Apparatus and method for approximating pixel data and reusing same in display

Abstract

An apparatus according to the present invention comprises: an approximation unit configured to identify a characteristic of pixel data in an image block of an arbitrary size or a characteristic of an area designated for the pixel data, and determine whether to approximate the pixel data on the basis of the identified characteristic, and further configured to apply, to arbitrary pixel data determined to be approximated in the image block, an approximation that creates missing data from the arbitrary pixel data by discarding a value of N bits (0 < N <= bit width of the pixel data) from lower bits; a buffer configured to store both the missing data to which the approximation has been applied and the pixel data to which the approximation has not been applied; and a control unit configured to receive the data stored in the buffer from a data supply unit that reads the data stored in the buffer according to a designated timing, and to visually express the data on a display panel.

Description

Device and method for approximating pixel data and reusing it on a display

[0001] The present invention relates to a device for visually displaying an image, and more specifically, to a device and method for reusing pixel data for display while selectively reducing information about pixel data constituting an image.

[0002] Display devices capable of outputting high-definition images are widely used, applied not only to TVs and monitors but also to portable devices and digital signage.

[0003] However, in the case of portable devices, power consumption is a major concern because they are supplied with necessary power by an onboard battery rather than being constantly connected to a power line. This is because the lower the power consumption required for operation, the longer the usage time of the portable device on a charged battery, thereby improving the user's convenience.

[0004] Display devices generally display images on the screen at a rate of 60 frames per second, or 60fps (frames per second) or higher. However, for display devices applied to portable devices, a method of reducing power consumption by lowering the frame rate below 60fps is sometimes applied because the proportion of power consumption in portable devices is high.

[0005] However, while this method can reduce power consumption in the display, it is not a desirable method because users experience image quality degradation as the frame rate is reduced, and a more desirable method for power consumption that takes these factors into account is needed.

[0006] One objective of the present invention is to provide an apparatus and method for reducing power consumption in image display in a manner that does not reveal degradation of image quality.

[0007] Another objective of the present invention is to provide an apparatus and method for reducing power consumed in recording and transmitting information by selectively reducing the amount of pixel information of an image input for screen display.

[0008] Another objective of the present invention is to provide an apparatus and method that enables an image with a reduced amount of pixel information to be displayed on a screen without any loss of image quality perceived by a human.

[0009] Another objective of the present invention is to provide an apparatus and method capable of preserving pixel information lost due to reduction of pixel information for an image by using residual information of pixel data used for screen display.

[0010] Another objective of the present invention is to provide an apparatus and method that enable the preservation of lost pixel information to be accurately achieved regardless of the characteristics of the display device or the operating environment.

[0011] The objectives of the present invention are not limited to those explicitly described above, but naturally include achieving effects that can be derived from the specific and exemplary description of the invention below.

[0012] An apparatus for reusing an input image on a display screen according to one aspect of the present invention is configured to identify the characteristics of pixel data within an image block of arbitrary size, or the characteristics of an area designated for said pixel data, and to determine whether to approximate said pixel data based on the identified characteristics; furthermore, for any pixel data for which approximation is determined within said image block, from the lower N ( 0 <N<=픽셀 데이터의 비트폭 )비트의 값을 버린 결손 데이터로 만드는 근사화를 적용하도록 구성된 근사화부와, 상기 근사화부에 의한 상기 근사화가 적용된 결손 데이터와 상기 근사화가 적용되지 않은 픽셀 데이터가 함께 저장되도록 구성된 버퍼와, 상기 버퍼에 저장된 데이터를 지정된 타이밍(timing)에 따라 읽어내는 데이터 공급부로부터 상기 버퍼에 저장된 데이터를 전송받아 구비된 디스플레이 패널 상에 시각적으로 발현되게 하는 제 1제어부를 포함하여 구성된다.

[0013] In one embodiment according to the present invention, the arbitrary pixel data to which the application of the approximation is determined has a brightness that is lower than or equal to a predetermined lower limit or greater than or equal to a predetermined upper limit. In this embodiment, if the degree to which the brightness of the arbitrary pixel data is separated from the lower limit or greater than the upper limit is large, the N is determined to be a larger value compared to when the degree of separation is small. Additionally, the approximation unit may determine the N for the second pixel data to be a larger value than the N for the first pixel data even when the degree to which the brightness of the first pixel data in the image block is separated from the lower limit and the degree to which the brightness of the second pixel data in the image block is separated from the upper limit are the same.

[0014] In one embodiment according to the present invention, the approximation unit determines whether to apply the approximation to any pixel data based on the distribution of hardness characteristic values ​​of pixel data within the designated area. Here, the hardness characteristic value refers to the ratio of the value reflecting the difference between pixel values ​​of any two pixel data to the value reflecting the spatial distance between those two pixel data. In this embodiment, the approximation unit compares the distribution of hardness characteristic values ​​of pixel data within the designated area with a plurality of reference distributions, identifies one reference distribution that is closest to the distribution of hardness characteristic values, and determines the value specified for the identified reference distribution as the N for the any pixel data. Additionally, if the distribution of hardness characteristic values ​​of pixel data within the designated area is skewed toward high hardness characteristic values, the N may be determined to be a larger value compared to when it is skewed toward low hardness characteristic values.

[0015] In one embodiment according to the present invention, the approximation unit applies the approximation to the arbitrary pixel data when the arbitrary pixel data differs from the average pixel value of the pixel data within the designated area by more than a predetermined reference value. In this embodiment, when the difference from the reference value of the pixel data based on the average pixel value is large, N may be determined to be a larger value compared to when it is small.

[0016] In another embodiment according to the present invention, the approximation unit calculates a weighted average by assigning weights to the number of bits to be discarded from the lower part obtained by at least two of the approximation methods presented in the above embodiments, based on the degree of brightness of the arbitrary pixel data and the degree corresponding to the edge, and determines N from the calculated weighted average.

[0017] In one embodiment according to the present invention, N may be determined by a calculation with a weight determined according to the difference between the contrast of the image block and a reference contrast determined corresponding to the spatial frequency of the image block.

[0018] In one embodiment according to the present invention, when the arbitrary pixel data falls within a specific pixel value range, the approximation unit changes the missing data created by the approximation into missing data obtained by performing an AND operation with a representative value determined for the specific pixel value range, thereby creating the final missing data. Here, the representative value is the pixel value with the largest number of lower consecutive zero bits among the pixel values ​​corresponding to the specific pixel value range. In this embodiment, if the approximation unit decides not to apply the approximation to a pixel data within the image, the pixel data is not subjected to an AND operation with the representative value even if it falls within the specific pixel value range.

[0019] In the preceding embodiment, the maximum value of the specific pixel value range may correspond to the largest value within the range of values ​​that the pixel data can have. Additionally, the specific pixel value range that is ANDed with a predetermined representative value may be set in two or more of the approximation units within the range of values ​​that the pixel data can have.

[0020] In one embodiment according to the present invention, the approximation unit can dynamically change the boundary of at least one side of the specific pixel value range based on the brightness of the surroundings of the device.

[0021] In one embodiment according to the present invention, the approximation unit applies an approximation to all pixel data within the image block, making them into defective data by discarding the N bit value from the lower. In this case, the approximation unit determines the number of bits to discard from the lower for each pixel data belonging to the image block based on the characteristics of the pixel data or the characteristics of the region designated for the pixel data, and selects one number as N from the set of bits to discard individually determined for the pixel data belonging to the image block. The number of bits selected as N to be applied to the approximation in the set may be the number of bits that occupies a maximum proportion greater than or equal to a predetermined reference value in the set, or the smallest number of bits greater than 0.

[0022] In another embodiment according to the present invention, N can be determined according to the degree of low contrast, where the contrast of the image block is low compared to a reference contrast determined in correspondence with the spatial frequency of the image block. In this embodiment, the approximation unit determines N as a value smaller than the value determined when the contrast of the image block is lower than the reference contrast when the contrast of the image block is higher than the reference contrast. Furthermore, regarding the reference contrast, it may be set differently for each spatial frequency of the image based on at least one of the brightness around the device and the distance from the device to the viewer. In this case, the reference contrast is set to have a relatively higher contrast when the brightness is brighter or the distance is greater. In addition, in this embodiment, the first bit number determined to be applied to the image block according to the degree of low contrast may be corrected based on the difference between the first bit number and the second bit number obtained by applying an approximation method not based on the reference contrast to the arbitrary pixel data, and the corrected first bit number may be applied as N to create missing data of the arbitrary pixel data.

[0023] In one embodiment according to the present invention, the approximation unit applies an approximation to all pixel data within the image block, making them into defective data by discarding the N bit value from the lower end, and after applying at least two different approximation methods to the image block, calculates the maximum number of approximation bits to discard that obtain an index greater than or equal to the quality index determined from the characteristics identified for the image, and selects the largest value among the calculated plurality of approximation bits as N.

[0024] In one embodiment according to the present invention, the approximation unit applies the approximation to the pixel data of the subsequent block when the block difference component obtained from the difference in pixel values ​​between at least some corresponding pixel data of the image block and the block that is temporally consecutive to the image block is less than or equal to a predetermined reference value.

[0025] In one embodiment according to the present invention, the approximation unit applies an approximation in which, for pixel data belonging to the image block, if the difference in pixel values ​​between the pixel data and the pixel data corresponding to the position of a block that is temporally consecutively following the image block is all below a reference value, the subsequent block is not stored in the buffer, thereby preventing it from being provided to the first control unit.

[0026] In one embodiment according to the present invention, the approximation unit applies an approximation to pixel data in the boundary zone of the image block such that, if the difference in pixel value between the pixel data and the pixel data corresponding to the position of a block that is temporally consecutively following the image block is less than or equal to a reference value for all pixel data in the boundary zone, the subsequent block is not stored in the buffer, thereby preventing it from being provided to the first control unit. Here, the boundary zone refers to the boundary of the image block or a boundary range having a bit width of 2 or more.

[0027] In one embodiment according to the present invention, the approximation unit applies an approximation in which, if the sum of the differences between each pixel data between the image block and a block that is temporally consecutive to the image block is less than or equal to a reference value, the subsequent block is not stored in the buffer, thereby preventing it from being provided to the first control unit.

[0028] In the aforementioned embodiments, multiple reference values ​​are specified, and the priority of application of the multiple reference values ​​in the comparison determining whether to apply the approximation to the pixel data may be changed. Additionally, these reference values ​​may be dynamically changed to relatively higher values ​​as the variability of the image increases.

[0029] In the aforementioned embodiments, the approximation unit provides the pixel data of the subsequent block to the first control unit by storing it in the buffer, even if the subsequent block is identified as not to be provided to the first control unit when comparing all or part of the pixel data of the image block with the subsequent block.

[0030] In the aforementioned apparatus and embodiments, the first control unit, in the transmitted data, has all of the data to be applied to any pixel unit among a series of pixel units belonging to a scan line on the display panel to be visually manifested at the current time, or K ( 0 from below <K<픽셀 데이터의 비트폭 )비트의 값이 버려진 부분 결손 데이터인 경우에는, 상기 임의의 화소 유닛에 현재 남아 있는 잔류값으로부터 상기 임의의 화소 유닛에 직전에 인가하였던 화소값을 구하고 그 구한 화소값 전부를, 또는 그 구한 화소값에서 상기 K비트에 대응되는 값을 부분 결손 데이터에 결합한 화소값을 상기 임의의 화소 유닛에 인가하도록 구성될 수 있다.

[0031] In one embodiment according to the present invention, the approximation of the image block or pixel data within the image block may be performed by the execution of a neural network that has been deep-learned on the approximation results of a plurality of images.

[0032] In one embodiment according to the present invention, the data supply unit transmits, regarding the approximation information for the pixel data to which the approximation is applied, a series of pixel data corresponding to any scan line on the display panel to which the pixel data to which the approximation is applied belongs, or between the transmission of an image frame containing the series of pixel data, or such transmission, to the first control unit. Here, the approximation information may be information indicating the number of bits discarded from the lower part of the pixel data to which the approximation is applied, or information indicating that the corresponding pixel data has been completely omitted due to the application of the approximation. In this embodiment, by transmitting the approximation information for the series of pixel data corresponding to the arbitrary scan line to the control unit based on the horizontal synchronization signal of the display panel, information regarding which image block the transmitted approximation information should be applied to may not be provided.

[0033] In one embodiment according to the present invention, the approximation unit stores only the remaining bits excluding the N bits for the missing data in the buffer, and also stores approximation information representing the number of the N bits in the buffer. In this case, the data supply unit may, for the missing data, refer to the approximation information for the missing data to fill the N bits from the lower end with 0 or 1 and transmit it to the first control unit. Alternatively, the first control unit may not drive the pixel unit at the position where the N bits of the missing data are to be expressed, thereby preventing them from being expressed as color or gradation.

[0034] In one embodiment according to the present invention, the image block of any size may be a block in which pixel data is modified by modeling the corresponding image block within the input image into a visual image perceived by the viewer in their current viewing environment.

[0035] According to another aspect of the present invention, an apparatus for rendering an input image on a screen comprises a display panel configured to include an array of pixel units that causes an applied pixel value to be visually expressed, and a second control unit configured to apply a pixel value corresponding to each pixel data to a pixel unit at a corresponding display position on the display panel so as to visually express the pixel data when pixel data belonging to an image frame is input. Furthermore, if there is missing data in which original information is lost within the image frame, the second control unit detects a residual value currently remaining in an arbitrary pixel unit corresponding to a position where the missing data is to be expressed on the display panel, constructs substitute data that preserves the lost information of the missing data based on the detected residual value, and then applies a pixel value corresponding to the substitute data to the arbitrary pixel unit.

[0036] In one embodiment according to the present invention, the second control unit preserves the lost information for the missing data in which the original information is completely lost or partially lost from the lower part, and configures substitute data to be applied to the pixel unit. In this embodiment, the second control unit may preserve the missing data in which the original information is partially lost from the lower part only when the upper part that is not lost in the missing data and the upper part of the pixel value immediately applied to the arbitrary pixel unit, which is obtained based on the detected residual value, are identical.

[0037] In another embodiment according to the present invention, the second control unit preserves the lost information only for the missing data in which all the original information has been lost, and configures substitute data to be applied to the pixel unit.

[0038] In one embodiment according to the present invention, the second control unit forms an electrical path with the arbitrary pixel unit, and detects the residual value based on a voltage in which the amount of charge currently remaining in the capacitor within the arbitrary pixel unit is balanced with another capacitor on the formed electrical path (e.g., a capacitor separately added to the electrical path or a parasitic capacitor existing in the electrical path) by charge sharing.

[0039] In one embodiment according to the present invention, the second control unit supplements the leakage amount during the time difference in the detected residual value based on the time difference between the time when a pixel value is immediately applied to the arbitrary pixel unit and the time when the missing data is preserved based on the residual value, and preserves the lost information of the missing data based on the residual value in which the leakage amount has been supplemented.

[0040] In one embodiment according to the present invention, the display panel is configured such that a source line commonly connected to apply a pixel value to a series of pixel units including the arbitrary pixel unit is composed of a plurality of line segments separated by a plurality of switches inserted and connected within the source line. In this embodiment, the second control unit detects the residual value of the arbitrary pixel unit after opening the switch connected to the line segment to which the arbitrary pixel unit belongs among the plurality of switches. In this embodiment, when there are a plurality of supply lines capable of applying a specific value to the arbitrary pixel unit, the signal line corresponds to the supply line with the smallest parasitic capacitance among the plurality of line segments.

[0041] In another embodiment according to the present invention, the display panel has a configuration in which a series of pixel units arranged in any column are distributed among a plurality of line segments arranged in a row while being electrically disconnected from each other, and a plurality of pixel units are commonly connected to each line segment.

[0042] In the aforementioned embodiments, each of the line segments is individually provided with signal lines that are directly connected from the outer edge of the display panel to the corresponding line segment. In this embodiment, each of the signal lines may be commonly used when applying a pixel value to pixel units that are commonly connected to the connected line segment and when detecting a residual value currently remaining in the pixel units.

[0043] In the aforementioned embodiments, the line segments have a structure in which they are individually connected to the lowest signal line links of signal line links having a tree structure, starting from one or more lines at the top from the outer edge of the display panel, and may be provided in such a form that a path selection switch is inserted at each point where the signal line links branch. In this embodiment, the signal line formed by the signal line links that are electrically connected to each other by the path selection of each path selection switch may be commonly used when applying a pixel value to pixel units that are commonly connected to the line segment to which the signal line is connected, and when detecting the residual value currently remaining in the pixel units.

[0044] In one embodiment according to the present invention, the display panel may further comprise a reference unit set in which a plurality of reference units having the same circuit structure as the pixel unit are arranged. In this embodiment, the second control unit may apply different pixel values ​​to the reference units within the reference unit set, and then compare the value detected at each reference unit of the reference unit set with the residual value, and use the pixel value applied to the reference unit in which the current value closest to the residual value is detected to preserve the lost information of the missing data.

[0045] In an embodiment in which the above reference unit set is provided, the display panel may individually provide a plurality of circuit lines having different electrical characteristics to transmit the current value of any reference unit within the reference unit set for comparison with the residual value. In this case, the second control unit selects one of the plurality of circuit lines according to which line segment the arbitrary pixel unit is connected to among a series of pixel units forming a column including the arbitrary pixel unit, and detects the current value of each reference unit through the selected circuit line and compares it with the residual value. Here, the series of pixel units are distributed and arranged among a plurality of line segments, and the pixel units distributed and arranged among each line segment are commonly connected to the corresponding line segment.

[0046] In addition, in an embodiment in which the above reference unit set is provided, at least one of the first signal line for detecting a current value from any reference unit within the above reference unit set and the second signal line for detecting the residual value of any pixel unit is inserted with a circuit that compensates for the difference in electrical characteristics between the first signal line and the second signal line up to a circuit in which the current value and the residual value are compared with each other.

[0047] In one embodiment according to the present invention, the reference unit set is configured such that the reference units are arranged in N1 x N2 (wherein N1 is any number greater than or equal to 2 and less than or equal to the number of possible pixel values, and N2 is any number greater than or equal to 1 and less than or equal to the number of pixel units on the scan line of the display panel), and the second control unit updates the reference unit set by recording the N1 different pixel values ​​identically to the N2 reference units for each pixel value whenever a pixel value is applied to each scan line of the display panel for the N1 x N2 reference units. In this embodiment, the N2 reference unit groups of the reference unit set are distributed and arranged at points that evenly partition the pixel columns of the display panel, and the control unit may be configured to include a plurality of comparison / estimation units that are provided at the top or bottom corresponding to each pixel column of the display panel, and compare the N1 current detection values ​​with the residual value detected from the pixel unit of the corresponding column and determine the nearest current value from the comparison. Each of these comparison / estimation units may be configured to compare the detected residual value with each pair of adjacent sizes for the N1 current detection values. Additionally, a value detected from any reference unit of the reference unit set is configured to be supplied to at least some of the plurality of comparison / estimation units by a pair of drivers, and the pair of drivers may be circuit-connected so that their outputs are commonly input to an equal number of comparison / estimation units.

[0048] According to another aspect of the present invention, an apparatus for rendering an input image on a screen comprises: a second display panel comprising an array of pixel units that cause an applied pixel value to be visually expressed, and a set of reference units arranged such that a plurality of reference units having the same circuit structure as the pixel unit are arranged therein; and a third control unit configured such that when pixel data belonging to an image frame is input, a pixel value corresponding to each pixel data is applied to a pixel unit at a corresponding display position on the display panel to cause it to be visually expressed. And, the third control unit detects a residual value currently remaining in any pixel unit on the second display panel based on a voltage created by charge redistribution to a capacitor on an electrical path formed for the any pixel unit, and estimates a pixel value immediately recorded in the any pixel unit through comparison with reference values ​​detected in a series of reference units within the reference unit set, and further, at least one of the two signal lines to which any one of the detected reference values ​​and the detected residual value are transmitted for comparison is configured to have a compensation circuit to make the electrical characteristics of the two signal lines the same.

[0049] In one embodiment according to the present invention, the second display panel has a supply line commonly connected to apply pixel values ​​to a series of pixel units including the arbitrary pixel unit, and the supply line is configured to consist of a plurality of line segments separated by a plurality of switches inserted and connected within the supply line, and the third control unit detects the residual value of the arbitrary pixel unit after opening the switch connected to the line segment to which the arbitrary pixel unit belongs among the plurality of switches.

[0050] According to another aspect of the present invention, a method for reusing an input image to display it on a screen comprises identifying the characteristics of pixel data within an image block of arbitrary size, or the characteristics of an area specified for said pixel data, and based on the identified characteristics, for said pixel data, from the lower N ( 0 <N<=픽셀 데이터의 비트폭 )비트의 값을 버린 결손 데이터로 만드는 근사화를 적용하는 단계와, 상기 이미지 내의 상기 근사화가 적용된 하나 이상의 결손 데이터와 상기 근사화가 적용되지 않은 픽셀 데이터들을 함께 버퍼에 저장하는 단계와, 상기 버퍼에 저장된 데이터를 지정된 타이밍에 따라 읽어내어 디스플레이 패널 상에 시각적으로 발현되게 하는 단계를 포함하여 이루어진다.

[0051] A method for reusing an input image to display it on a screen, according to another aspect of the present invention, comprises: a first step of receiving pixel data belonging to an image frame; and

[0052] The method comprises a second step of applying a pixel value corresponding to each of the received pixel data to a pixel unit at a corresponding display position on a display panel to make it visually manifest. Here, the second step includes, if there is missing data in which original information is lost within the image frame, detecting a residual value currently remaining in an arbitrary pixel unit corresponding to a position where the missing data is to be displayed on the display panel, constructing substitute data that preserves the lost information of the missing data based on the detected residual value, and then applying a pixel value corresponding to the substitute data to the arbitrary pixel unit.

[0053] In the aforementioned apparatus and method and various embodiments, the designated area may be a block having a certain size centered on the arbitrary pixel data.

[0054] An apparatus for approximating and reusing pixel data for display on a screen, according to the present invention described above or at least one embodiment of the present invention described below in detail with the attached drawings, can reduce the power consumed in recording and transmitting information contained in pixel data by reducing the amount of information recorded when temporarily storing or updating pixel data of a received image in memory for display. That is, the present invention can provide a display device with higher energy efficiency.

[0055] In addition, in one embodiment according to the present invention, even when the amount of information regarding pixel data included in an image is reduced, the information of the pixel data lost due to the reduction of information is preserved by using the residual value of the pixel data previously used for screen display, and then displayed on the screen, thereby preventing visually perceptible degradation of image quality despite the reduction of pixel data information for energy saving.

[0056] FIG. 1 is an example of the configuration of a display device according to an embodiment of the present invention, which selectively approximates pixel data within a received image and renders an image including the approximated pixel data on a screen to make it visually displayed.

[0057] FIG. 2 schematically illustrates pixel units arranged in a grid shape on a display panel, and

[0058] FIG. 3 illustrates a configuration in which, according to one embodiment of the present invention, a source line is divided into a plurality of line segments by inserting a plurality of switch elements in series into a single source line, and a recovery line is connected to each line segment.

[0059] FIG. 4 is a flowchart of a method for selectively approximating pixel data in a received image according to an embodiment of the present invention, and

[0060] FIG. 5a schematically illustrates a process of selectively applying an approximation that discards some LSB values ​​for arbitrary pixel data according to an embodiment of the present invention.

[0061] FIG. 5b exemplarily illustrates a zero packing method based on a representative value applied to selectively discard lower bits after approximating pixel data for a specific pixel value range, according to an embodiment of the present invention.

[0062] FIG. 5c schematically illustrates a method for determining the degree of approximation for an image block of an image based on a contrast sensitivity function according to another embodiment of the present invention, and

[0063] FIG. 6 schematically illustrates a process of approximating an arbitrary block of an image based on the correlation between temporally consecutive image blocks according to an embodiment of the present invention, and

[0064] FIGS. 7a and 7b schematically illustrate a configuration and process for supplementing input pixel data and applying it to a corresponding pixel unit when there are approximated bits, according to an embodiment of the present invention.

[0065] FIGS. 8a to 8c each show an equivalent circuit for a configuration for detecting a voltage remaining in an arbitrary pixel unit on a display panel according to an embodiment of the present invention, and

[0066] FIG. 9 illustrates a structure in which, according to another embodiment of the present invention, recovery lines connected to each of a plurality of line segments corresponding to a single source line are configured as circuit line links of a hierarchical structure.

[0067] FIG. 10 illustrates a configuration in which, according to another embodiment of the present invention, pixel units forming a column are distributed into a plurality of electrically pre-disconnected line segments and connected in common to each line segment, and signal lines for recording and reading pixel values ​​are connected to each line segment.

[0068] FIG. 11a illustrates the configuration of a reference pixel unit set provided in a display panel to accurately estimate a previously applied pixel value from a residual value of a pixel unit regardless of variability in the operating environment of the display panel, according to one embodiment of the present invention.

[0069] FIG. 11b is a drawing showing an embodiment of the present invention regarding the configuration of a pixel unit of a display panel and the connection of a signal line for recovering pixel values ​​accordingly, and

[0070] FIGS. 12a and 12b respectively illustrate a part of the internal configuration of a comparison estimation module for comparing a residual value detected from a pixel unit with the current value of reference units according to embodiments of the present invention, and

[0071] FIG. 12c illustrates the configuration of a comparison unit according to embodiments of the present invention, which enables determining the pixel value immediately applied to the pixel unit where the residual value was detected by comparing the residual value of the pixel unit and the current values ​​of the reference unit within the comparison estimation module of FIG. 12a and 12b.

[0072] FIG. 13a is a block diagram schematically illustrating the configuration when a comparison estimation module is implemented so that all pixel units belonging to a single scan line commonly use a reference unit set, according to an embodiment of the present invention.

[0073] FIG. 13b conceptually illustrates the layout of a comparison estimation module on a display panel according to an embodiment of the present invention, and

[0074] FIG. 14 illustrates the configuration of a line characteristic compensation unit for adaptively changing the circuit characteristics of a drawing line for detecting a series of voltages recorded in a reference pixel unit set, according to one embodiment of the present invention.

[0075] Hereinafter, various embodiments according to the present invention will be described in detail with reference to the attached drawings.

[0076] In the following description of embodiments according to the present invention and the accompanying drawings, identical numbers refer to identical components unless there are special circumstances. Of course, for the convenience of explanation and to aid understanding, identical components may be assigned different numbers as necessary.

[0077] FIG. 1 is an example of the configuration of a display device according to an embodiment of the present invention, which selectively approximates pixel data within a received image and renders an image containing the approximated pixel data onto a screen to make it visually displayed.

[0078] The display device (100) illustrated in FIG. 1 may be included in TVs, monitors, laptops, particularly portable communication devices, such as smartphones, as well as any product to which a means of visually displaying images to a user is added. Accordingly, the scope of the claims of the present invention should not be excluded on the grounds that the commonly referred to name of the product or goods is different.

[0079] Looking at the configuration of the device (100) exemplified in FIG. 1, for images received sequentially in time, e.g., video frames, there is an approximation unit (101) that selectively applies an approximation of pixel data within each image, a frame buffer (102) in which image data including the selectively approximated pixel data is temporarily stored, and pixel units (CC) that form a grid-shaped array on the display panel (120) as shown in FIG. 2, each having a voltage corresponding to each pixel data. i,j Cell capacitor (C) of , i=1,2,..,n,..,SNum, j=1,2,..,m,..,M) C A display panel (120) that visually expresses a color (or gradation) corresponding to the charged voltage by charging it, and each source line (S i While driving each gate line (G , i=1,2,..,n,..,SNum) with a voltage corresponding to the pixel data j The system comprises a display control unit (110) that visually displays an image on the display panel (120) (hereinafter abbreviated as 'panel') by controlling signals of j=1,2,..,m,..,M), and a data supply unit (103) that transmits pixel data of an image, some of which has information reduced by approximation and is stored in the frame buffer (102), in a data format required by the display control unit (110).

[0080] The display control unit (110) includes an internal frame buffer that separately stores pixel data supplied from the data supply unit (103), controls the timing for displaying the pixel data supplied from the data supply unit (103) and stored in the internal frame buffer on the display panel (120), and also includes a necessary interface. The timing at which the data supply unit (103) supplies pixel data to the display control unit (110) is also determined by the timing controlled by the display control unit (110).

[0081] In one embodiment according to the present invention, the panel (120) is an arbitrary data line (hereinafter referred to as a 'source line' in this specification, meaning a line on which a source of data to be visually expressed is carried) on which a voltage corresponding to pixel data is carried, as illustrated in FIG. 3. n Regarding ), pixel units (pxU) commonly connected to the source line are grouped into multiple units (pxG) as units, and a switch (SW), such as a transistor, is used. n_i It is possible to have a circuit structure in which , i=1,2, ) are inserted in series. And, each switch (SW) inserted and provided in the source line in such a way can be configured. n_i At the gate of ), an opening / closing line (scL) through which a signal for controlling the On / Off of the switch is transmitted n_S[i] ) is connected, and those switches (SW n_i Each source line segment separated by ) has a recovery line (rL n_S[i] In a manner where ) are connected, the recovery lines are provided in parallel with respect to the source line segments.

[0082] In addition, in this embodiment, the display control unit (110) is the opening / closing line (scL n_S[i] ) and the above recovery lines (rL n_S[i]By separately providing a configuration that is circuitously connected to each of the ) to visually represent an image on the panel (120) source lines (S i ) and gate line(G j In addition to the general signal applied to ), through that separate circuit configuration, a necessary signal (c110) is applied or a desired signal (r110) is read. This will be explained in detail later.

[0083] Hereinafter, the operation of selectively approximating each pixel data included in a received image and rendering the approximated pixel data on a screen to make it visually represented, which is performed by a display device (100) having the configuration exemplified in FIG. 1, will be explained in detail through specific examples.

[0084] First, we will explain a method for selectively approximating pixel data contained in an input image.

[0085] FIG. 4 is a flowchart of a method for selectively approximating pixel data in an input image according to an embodiment of the present invention.

[0086] According to the flowchart illustrated in FIG. 4, the approximation unit (101) selects one pixel data within an image according to the order of the raster scan method for an input image frame, i.e., an image (S41). In a preferred embodiment according to the present invention, the input image may be divided into image blocks of arbitrary size, and pixel data within the corresponding image block may be selected according to the order of the raster scan method, with each image block as a unit.

[0087] Then, determine whether the brightness characteristics of the selected pixel data correspond to the approximation target. Determining whether the brightness characteristics correspond to the approximation target is done by comparing the value represented by the selected pixel data with a predetermined upper limit, which is a reference value for brightness, and a lower limit, which is a reference value for darkness.

[0088] The value represented by the pixel data may be a digital value of the pixel data, or a luminance value based on the three primary color pixel data having the pixel data as a subpixel—where the values ​​of each subpixel are Rv, Gv, and Bv, for example, a digital value obtained by the formula Y = 0.299 χRv + 0.587 χGv + 0.144 χBv. In this specification, this digital value is referred to as a 'pixel value' or a 'gradation value'. Additionally, in this specification, depending on the case, the terms 'pixel value' or 'gradation value' may refer to a value in which the corresponding digital value is converted into an analog voltage for visual display on a screen.

[0089] The approximation unit (101) checks whether the pixel value of the selected pixel data is a value greater than or equal to the upper limit (i.e., a pixel value brighter than or equal to the upper limit) or a value less than or equal to the lower limit (i.e., a pixel value darker than or equal to the lower limit) (S42). The upper limit and the lower limit may be 240 and 16, respectively, assuming the bit width of the pixel data is 8 bits.

[0090] If the pixel value is greater than or equal to the upper limit or less than or equal to the lower limit, the approximation unit (101) applies an approximation that discards some LSB values ​​for the corresponding pixel data, and the approximation strength, that is, the number of LSB values ​​to discard, is determined based on the difference between the pixel value and the upper limit or lower limit (S43).

[0091] FIG. 5a schematically illustrates such an approximation process, showing that when the corresponding pixel data (510) becomes the target of approximation (50) such that it corresponds to a pixel value that exceeds the upper or lower limit, some bits of the corresponding pixel data are discarded (520). In this specification, the number of LSBs to be discarded refers to the number of lower bits among the bits (b7 to b0) corresponding to the pixel. For example, if the number of LSBs to be discarded is 3, it means that b0, b1, and b2 are discarded.

[0092] The example in FIG. 5a shows that the number of discarded LSBs, i.e., the approximation strength, is selected from 1, 3, or 5, and this strength (number of discarded LSBs) is determined to be stronger as the difference between the pixel value of the corresponding pixel data and the upper or lower limit is greater (i.e., the number of discarded LSBs is determined to be a larger value). Of course, the number of discarded LSBs may be selected and applied differently from the example, for example, one of the cases of 2, 4, and 6, or one of the cases of 1, 4, and 7.

[0093] In one embodiment according to the present invention, different approximation intensities may be applied to cases where the pixel value of the pixel data is separated toward the bright side (larger value side) and cases where it is separated toward the dark side (smaller value side). That is, in cases where the difference in pixel value from the upper limit and the difference in pixel value from the lower limit are the same, the approximation intensity may be applied more strongly to the difference from the upper limit. For example, for a first pixel data where the pixel value is pv larger than the upper limit and a second pixel data where the pixel value is pv smaller than the lower limit, it may be determined that 5 bits of LSB are discarded for the first pixel data and 3 bits of LSB are discarded for the second pixel data.

[0094] Meanwhile, the approximation unit (101) adds information indicating the number of LSBs discarded (hereinafter referred to as 'approximation strength information') to the pixel data to which approximation has been applied, that is, the approximated pixel data. In the example of FIG. 5a, 2 bits are allocated and used for the approximation strength information (521), and each value (0, 1, 2, 3) possible with 2 bits is designated to indicate that the number of LSBs discarded is 0, 1, 3, and 5, respectively.

[0095] The approximation unit (101) writes bits of pixel data excluding discarded LSBs and bits of approximation intensity information (c1c0) to the frame buffer (102) in the manner exemplified in FIG. 5a (S60).

[0096] Meanwhile, if the current pixel data has a pixel value within the range of the upper and lower limits, the approximation unit (101) checks whether the pixel data corresponds to a contour, that is, an edge (S44). Whether the pixel data corresponds to an edge can be checked by applying any one of the generally known methods, such as the Sobel edge detection method or the Canny edge detection method.

[0097] If, in the edge detection method applied to the pixel data, an edge component (a value indicating the degree of edge obtained from the applied edge detection algorithm) that can be identified as an edge is detected, the approximation unit (101) designates a pixel block of a certain size, for example, 5x5 or 7x7, centered on the pixel data. At this time, the pixel block designated is not limited by the image block to which the pixel data belongs. That is, pixel data belonging to an image block adjacent to the current image block may be included and designated as the pixel block for the pixel data.

[0098] When a pixel block is specified, the distribution characteristics of the inclination are calculated for that pixel block (S45). Here, 'inclination' refers to the ratio of the value reflecting the difference in pixel values ​​to the value reflecting the spatial distance between pixel data. The simplest method to calculate the individual inclination between two given pixel data is to divide the difference in pixel values ​​between the two pixel data (this difference always has a positive value) by the distance between the two pixel data. The inclination distribution characteristics for a pixel block may be a characteristic represented by the number of individual inclinations belonging to each inclination range when the individual inclinations of the pixel data belonging to the pixel block are classified into a number of defined inclination ranges.

[0099] The approximation unit (101) determines the approximation strength for the pixel data according to the distribution characteristics when the hardness distribution characteristics for the pixel block are obtained. For example, if the obtained distribution characteristics show a shape skewed toward high hardness, a relatively stronger approximation strength is applied compared to a shape skewed toward low hardness. That is, the number of LSBs to be discarded relatively for the pixel data is determined to be a larger value.

[0100] In another embodiment according to the present invention, the obtained hardness distribution characteristic may be compared with a plurality of previously set reference distribution characteristics, and a specified approximation strength may be applied to the reference distribution characteristic that represents the closest distribution characteristic. In this embodiment, the approximation unit (101) compares the obtained hardness distribution characteristic with a plurality of previously set reference distribution characteristics to find the closest reference distribution characteristic. The proximity between distribution characteristics may be determined based on the standard deviation for the convenience of calculation. The smaller the difference in standard deviation between the reference distribution characteristic and the corresponding distribution characteristic, the more similar the distribution characteristics are determined to be. The difference between the reference distribution characteristic and the corresponding distribution characteristic can be called the degree of proximity. The reference distribution characteristic with the smallest degree of proximity is determined to be the closest to the hardness distribution characteristic of the corresponding pixel block.

[0101] When the closest reference distribution characteristic is found in this way, the approximation unit (101) determines the approximation strength to be applied to the corresponding pixel data using a setting value specified for the reference distribution characteristic. That is, if the value specified for the reference distribution characteristic is N, the number of LSBs to discard for the corresponding pixel data is determined to be N (S46).

[0102] The values ​​assigned to the reference distribution characteristics that are set in the approximation unit (101) may be 0. If it is determined that the longitude distribution characteristic of the corresponding pixel block is closest to the reference distribution characteristic to which 0 is assigned, approximation is not applied to the corresponding pixel data, which means that the corresponding pixel data is preserved intact without pixel information reduction and used for post-processing.

[0103] In another embodiment according to the present invention, the distribution characteristic of hardness to which approximation is not applied (to which 0 is assigned as a designated value) may not be pre-set as a reference distribution characteristic in the approximation unit (101). In this embodiment, when finding the reference distribution characteristic that is closest to the hardness distribution characteristic of the corresponding pixel block, if the degree of closestity (the minimum degree of closestity) is not less than or equal to a preset limit, it is determined that there is no reference distribution characteristic similar to the hardness distribution of the corresponding pixel block, and the approximation strength for the corresponding pixel data is determined to be 0. That is, it is determined not to approximate the corresponding pixel data.

[0104] Meanwhile, if the edge component of the pixel data is not considered an edge, the approximation unit (101) specifies a pixel block of a certain size centered on the pixel data as described above and calculates the average value of the pixel values ​​for the pixel block. Then, by comparing the difference between the calculated average value and the pixel value of the corresponding pixel data with a preset reference value, it determines whether the corresponding pixel data is a target for approximation (S51).

[0105] The approximation unit (101) determines a pixel data as an approximation target if the difference between the pixel value of the pixel data and the average value of the pixel value of the corresponding pixel block is greater than a preset threshold (meaning that the pixel value is a value that stands out more than the threshold set in the corresponding pixel block), and determines a larger number of LSBs to discard from the corresponding pixel data as the degree to which the difference exceeds the threshold (S52). Then, when the number of LSBs to discard, i.e., the approximation strength, is determined, the bits of the pixel data excluding the determined number of LSBs and the approximation strength information (521) are recorded in the frame buffer (102) as described above (S60).

[0106] If the difference between the pixel value of the pixel data and the average value of the corresponding pixel block is less than or equal to the reference value, the approximation unit (101) determines that the number of LSBs to discard for the corresponding pixel data is 0 (S53), thereby not approximating the pixel data.

[0107] In one embodiment according to the present invention, the reference value may be set based on the standard deviation representing the pixel value distribution characteristics of pixel data within the corresponding pixel block. For example, the reference value is set to a value obtained by multiplying the pixel value standard deviation of the corresponding pixel block by an adjustment constant that is adaptively determined according to the required display quality.

[0108] In another embodiment according to the present invention, the reference value may be determined based on a characteristic value representing the inclination of the corresponding pixel block. Here, the characteristic value representing the inclination may be the average of the individual inclinations of the pixel data within the block.

[0109] After determining the number of LSBs to discard, i.e., the approximation strength, to be 0 or some value, the approximation unit (101) writes the bits excluding the LSBs to be discarded from the corresponding pixel data to the frame buffer (102), along with the bits (521, 531) of approximation strength information indicating the value, as in FIG. 5a (S60).

[0110] In another embodiment according to the present invention, approximate intensity information may not be added to the approximated pixel data. In this embodiment, the approximated pixel data, i.e., pixel data where some LSB bits are discarded, is displayed on the screen as the corresponding color or gradation. That is, in this embodiment, the operation of preserving the value of the discarded LSB bits based on the residual voltage of the corresponding pixel unit, which will be described later, is not applied.

[0111] As described above, if pixel data within the received image or within each divided image block is selectively approximated and recorded in memory, the number of bits to be recorded is statistically reduced, thereby reducing the amount of memory occupied and the electrical energy required to record data in memory.

[0112] For example, when an experiment was conducted on arbitrary images by applying a method in which 1, 3, or 5 bits are selectively discarded for pixel data with a bit width of 8 and 2 bits are added as approximate intensity information, it was confirmed that on average about 6 to 7 bits are recorded for pixel data with a bit width of 8 (10-bit data including approximate intensity information).

[0113] In the aforementioned embodiment, when selectively approximating pixel data, a single method suitable for the characteristics of the pixel data was applied as exemplified in FIG. 4. However, in another embodiment according to the present invention, the aforementioned approximation method may be applied in combination to a single pixel data. The present embodiment will be described in more detail below.

[0114] In the present embodiment, which applies a combination of approximations to a single pixel data, if the pixel data corresponds to an edge, a selective approximation method based on the pixel value (a method in which step S43 is selectively applied by step S42, hereinafter referred to as the 'first method') and a selective approximation method based on the hardness distribution characteristics of a pixel block of a certain size centered on the pixel data (a method in which steps S45 and S46 are selectively applied, hereinafter referred to as the 'second method') are applied, and if the pixel data does not correspond to an edge, the first method and a selective approximation method based on the difference between the pixel data and the average pixel value of the pixel block (a method in which step S52 is selectively applied by step S51, hereinafter referred to as the 'third method') are applied.

[0115] And, when it is determined that a single pixel data is approximated by both approximation methods, a weight is applied to the approximation strength determined by each method to determine the final approximation strength. For example, if the given pixel data corresponds to an edge and the first method and the second method are applied together, and as a result the number of LSB bits to be discarded in each are determined to be N1 and N2, the approximation unit (101) determines the final number of bits to be discarded N from the weighted average value in which the number of bits to be discarded determined by each method is reflected in the ratio of weights α1 and α2 (where α1+α2=1). f ( = rounding(α1·N1+α2·N2) ) is determined and the corresponding pixel data is approximated as described above.

[0116] The weights α1 and α2 to be applied to each method can be adaptively determined according to the attributes of the pixel data. For example, if the edge component of the pixel data is large, the weight α2 applied to the result of the second method is set to a larger value and used accordingly.

[0117] Even when the above-mentioned first and third methods are applied simultaneously, as previously stated, the weights (β1 and) assigned to each method The weighted average calculated by applying β3 (β1+β3=1) is determined as the final number of LSB bits to discard. At this time, each weight β1 and β3 can also be adaptively determined and used according to the attributes of the pixel data. For example, even if the degree to which the pixel data does not correspond to an edge is the same, if the brightness of the pixel data is relatively larger, β1 applied to the result of the first method can be set to a larger value and used compared to when the pixel data is not.

[0118] In another embodiment according to the present invention, the final approximation strength may be determined in a composite manner by applying all three of the aforementioned approximation methods to a single pixel data. In this embodiment, the number of LSB bits to be discarded is determined by applying weights to each of the first, second, and third methods according to the degree to which the corresponding pixel data corresponds to an edge, that is, the degree of deviation between the edge component and the reference value of brightness. That is, when the approximation strengths obtained by applying the first, second, and third methods to the corresponding pixel data are denoted as N1, N2, and N3, respectively, the final approximation strength N is obtained from the weighted average value calculated by reflecting each of them in proportions of the weights γ1, γ2, and γ3 (where γ1+γ2+γ3=1, γ1, γ2, γ3>0). f Determine ( = rounding(γ1·N1+γ2·N2+ γ3·N3) ) and apply approximation to the corresponding pixel data.

[0119] Of course, the weight γ1 applied to the approximation strength by the first method is determined to be a large value depending on the degree of difference if the brightness of the corresponding pixel data is brighter or darker than the reference value, and the weight γ2 applied to the approximation strength by the second method is determined to be a large value if the edge component of the corresponding pixel data is large. And, if the brightness of the corresponding pixel data is within the reference value and the edge component is also weak, γ3 applied to the approximation strength by the third method is determined to be a large value and used to calculate the final approximation strength.

[0120] In one embodiment according to the present invention, representative value masking may be applied to any pixel data in which the approximation strength is determined according to the various approximation methods described above. Representative value masking is applied when the pixel data has a value that belongs to at least one interval determined for a part of the range that the pixel value can have.

[0121] FIG. 5b schematically illustrates the representative value masking according to the present embodiment, using the section with the highest luminance (pVS) in at least one section within the range designated for the application of the representative value masking as an example. The numerical values ​​specifically shown in FIG. 5b are intended only to aid in understanding the representative value masking and are not necessarily bound when the representative value masking according to the present embodiment is applied.

[0122] The range designated for applying representative value masking corresponds to a range of pixel values ​​where the human eye becomes relatively insensitive to detecting changes in brightness due to changes in pixel values, and the range may be designated as a single range or divided into multiple ranges.

[0123] Regarding the section where the representative value masking is applied, as exemplified in FIG. 5b, the representative value (541) is determined, and in the corresponding section (pVS), the pixel value (541) with the most consecutive zeros in the lower bit is determined as the representative value. If there is a section other than the exemplified section (pVS), the representative value is determined for that section as well according to the same principle.

[0124] At least one representative value masking application section designated within the range of all pixel values ​​and the representative value in that section are set in the approximation section (101), and the set representative value functions as a zero packer for the pixel values ​​belonging to that section as described below.

[0125] The approximation unit (101) fills the pixel data with zeros equal to the number of consecutive zeros of the lower bits of the representative value (541) of the corresponding section when the pixel data, whose approximation strength is determined according to one of the various embodiments described above, falls within a predetermined representative value masking application section. This operation can be performed by a bit AND operation that masks the representative value for the pixel data.

[0126] In a specific example shown numerically in FIG. 5b, the representative value (541) is masked onto the approximated data (lower 2 bits filled with 0) of pixel data (5511) with an approximation strength determined to be 2, so that the upper 4 bits retain the original value of the pixel data and the lower 4 bits are filled with 0, and the pixel data (5512) is output as the final pixel data.

[0127] If the determined approximation strength for the pixel data is greater than the number of consecutive zeros in the lower part of the representative value (5521), then, by approximation, more bits than the lower bits that are to be filled with zeros by masking the representative value (541) are already filled with zeros, so the value is not affected by masking with the representative value, and the approximated pixel data is output as the final pixel data (5522).

[0128] Meanwhile, the approximation unit (101) does not apply representative value masking to pixel data even if the pixel data of a pixel value belonging to the representative value masking application interval is 0, that is, if the approximation strength determined by the aforementioned approximation is 0, i.e., if the pixel data (5531) is determined not to be approximated by even one bit, then the representative value masking is not applied to that pixel data. In other words, the input pixel data is output as is as the final pixel data (5532).

[0129] In one embodiment according to the present invention, the representative value masking application section is not fixed and can be dynamically changed and applied according to viewing environment variables such as the ambient brightness of the display panel or the distance from the panel to the viewer. Referring to FIG. 5b as an example, the section lower limit (S LB ) can be changed to a higher or lower value from a set default value depending on the ambient brightness of the panel and applied to the representative value masking operation. Specifically, if it is higher than the default illuminance, the lower limit of the range (S LB ) is adjusted to a higher value, and to a lower value if it is low. This is because, as the surrounding environment becomes brighter, changes in brightness can be detected more effectively visually.

[0130] If the entire range to which representative value masking is applied is divided into multiple sections, the lower limit of the entire range to which representative value masking is applied is changed to a higher or lower value according to the change in illumination, and then the upper and lower limits of each section, excluding the upper limit of the entire range, are changed according to the rate of change of that lower limit. The approximation of pixel data described so far was based on an embodiment in which approximation intensity is applied individually to pixel data belonging to an image block of arbitrary size. That is, it was an embodiment in which approximation intensity information determined for each pixel data is provided to the display control unit (110) by the data supply unit (103) together with the corresponding pixel data. In these embodiments, since approximation intensity information must be recorded and provided for each pixel data, the effect of reducing the amount of data in memory recording and transmission by approximation may be limited.

[0131] Accordingly, in another embodiment according to the present invention, the approximation strength is applied equally to image blocks of arbitrary size. In this embodiment, since approximation strength information is not required individually for each pixel data but only one approximation strength information is required for each image block, it can bring about a significant reduction in the amount of data in memory recording and transmission through approximation. The embodiments described below will be explained in more detail.

[0132] In this embodiment, for each pixel data belonging to a given image block, one of the methods 1, 2, and 3 is applied, or two or more methods are combined to determine the approximation strength. Then, one approximation strength is selected from the set of approximation strengths of each pixel data determined in this way.

[0133] As a method of selecting one approximation strength from a set of approximation strengths, the approximation strength that occupies the largest number in the set is selected. For example, if, in the set of approximation strengths determined for pixel data belonging to a given image block, approximation strength 0 (no approximation) occupies 10%, 1 occupies 15%, 3 occupies 70%, and 5 occupies 5%, the selected approximation strength becomes 3. Then, approximation is applied to all pixel data belonging to the image block using the approximation strength selected in this way. That is, 3 bits of LSB are discarded for all pixel data of the image block. Then, for the image block, only one approximation strength information, namely the numerical value representing the selected 3, is subsequently provided to the display control unit (110).

[0134] Another method for selecting a single approximation strength from a set of approximation strengths is to minimize approximation by considering the image quality. In this method, the lowest approximation strength (excluding the approximation strength of 0) is selected from the set of approximation strengths; for example, if the approximation strengths are distributed as 1, 3, and 5, 1 is selected as the approximation strength, and if they are distributed as 3, 4, and 6, 3 is selected as the approximation strength. Then, the approximation strength selected in this way is applied to the approximation of all pixel data belonging to the corresponding image block.

[0135] In this specification, the description of applying the first to third methods to an image block to approximate it should be understood to mean that a single approximation strength selected according to one of the methods previously exemplified is applied to the entire block, based on the approximation strengths obtained for each pixel data according to the method.

[0136] In another embodiment according to the present invention, the approximation strength for an image block is determined based on a contrast sensitivity function (CSF). FIG. 5c schematically illustrates a method for determining approximation strength according to the present embodiment, wherein the approximation unit (101) has a contrast sensitivity function (CSF) as exemplified in FIG. 5c. i A data set representing , i=1,2,..) is recorded in advance.

[0137] The contrast sensitivity function is a graph representing the boundary line at which the human eye can distinguish differences in brightness between pixels on a plane where image contrast and spatial frequency are the coordinate axes. Based on this boundary line, if the contrast becomes lower than it (upper part of the graph), it means that visual differences in brightness cannot be detected or become difficult to perceive. Since it becomes more difficult for the human eye to detect differences in brightness between pixel values ​​even at the same contrast level when the image's spatial frequency increases, the contrast sensitivity function takes the form of decreasing sharply as the spatial frequency increases or decreases, as exemplified.

[0138] In this embodiment, in which the approximation strength is determined based on the contrast sensitivity function, the approximation unit (101) first obtains the contrast and spatial frequency for each image block of the image. Once the contrast and spatial frequency are obtained (IB1, IB2), the same spatial frequency (SF1) of the previously recorded contrast sensitivity function (CSF1) is obtained. P1 ,SF P2 Contrast (C) for ) Fk_i Difference value with ) { D CS_P1 , D CS_P2 (=C Pi - C F1_i )} is calculated, and the difference value is applied to a predetermined approximation function to determine the approximation strength (S50). The approximation function applied here (F(D CS)) can vary. Of course, this function may be in the form of a table that divides the entire range of contrast difference into several intervals and maps the approximation strength applied to each interval.

[0139] Approximation function(F(D CS )) is the contrast difference with the contrast sensitivity function (=C Pi - C F1_i If the rule that the larger ) is, the larger the approximation strength of the larger value is determined is observed, it can be applied as an embodiment of the present invention regardless of the form or format of the function. If the contrast difference calculated as above is negative (IB2), no approximation is performed (approximation strength = 0), or its magnitude (=|C Pi - C F1_i If |) is within a certain range, a pre-specified minimum approximation strength can be applied.

[0140] In an embodiment in which the approximation strength is determined based on a contrast sensitivity function, the contrast sensitivity function (CSF) that is pre-recorded in the approximation unit (101) i There may be multiple contrast sensitivity functions. These multiple contrast sensitivity functions may be determined for segments divided by the range that viewing environment variables, such as the ambient brightness of the display panel (120) or the distance from the panel to the viewer, may have. Thus, the multiple contrast sensitivity functions (CSF) provided in the approximation unit (101) i , i=1,2,..) is a set of functions of a characteristic (501) in which boundary contrast relative to the same spatial frequency is lowered for brighter ambient brightness or a longer viewing distance.

[0141] According to the present embodiment, when the approximation unit (101) applies approximation to an input image, it first checks the ambient brightness of the display panel (120) and the viewing distance of the viewer, and then selects a contrast sensitivity function determined for the ambient brightness and viewing distance from a plurality of contrast sensitivity functions, and determines the approximation intensity for each image block based on the selected contrast sensitivity function as described with reference to FIG. 5c. In addition, in the present embodiment, a detection means (e.g., a sensor, etc.) necessary for checking the ambient brightness and viewing distance of the display panel (120) is provided in the display device (100) to provide the detected information to the approximation unit (101). The viewing distance information may be set in the display device (100) by a separate user interface means.

[0142] In one embodiment according to the present invention, a 4-method for determining the approximation strength for an image block based on the contrast, spatial frequency, and contrast sensitivity function of the image block may be used to determine the weights of the approximation strengths individually determined for pixel data by the previously described 1 to 3 methods. In this embodiment, the contrast difference (=C) determined by the 4-method Pi - C F1_i If ) is greater than 0, a weight greater than 1 and proportional to the magnitude of the difference; if ) is less than 0, a weight less than 1 and proportional to the magnitude of the difference (|C Pi - C F1_i A contrast difference-weight conversion function can be applied to obtain weights inversely proportional to |). For example, for an arbitrary image block, the weight obtained by applying the above 4 method is W CSF Then, for each pixel data of the corresponding block, the weight W is applied to the approximation strength obtained by any one of the methods 1 to 3 or a combination thereof. CSFThe result of multiplying is corrected by the approximation strength, and the corrected strength is applied to the corresponding pixel data as the approximation strength.

[0143] Correction for approximation strength may be applied in reverse. That is, the block approximation strength for any image block is determined by the above-mentioned method 4 and applied to the entire block, but for each pixel data, the correction is applied by adding or subtracting a correction value equal to the correction ratio of the difference between the approximation strength determined by any one of the above-mentioned methods 1 to 3 or a combination thereof and the block approximation strength, and then the corrected approximation strength can be applied individually. For example, assuming the above correction ratio is 0.5, if the approximation strength determined for any pixel data is 3 and the block approximation strength determined by the above-mentioned method 4 for the image block to which that pixel data belongs is 1, then for that pixel data, the correction value becomes 1 ( = 0.5 × (3 - 1) ), and a final approximation strength of 2 ( = 1 + 1 ) can be applied. Of course, when the approximation strength obtained for that pixel data is smaller than the block approximation strength obtained for that block, the correction value can be negative.

[0144] In another embodiment according to the present invention, the approximation strength to be applied to the entire image block can be determined as a maximum value based on image quality. In this embodiment, the characteristics of the input image or image block are identified first, and an image quality index stored in advance for those characteristics is used.

[0145] The approximation unit (101) applies a predetermined algorithm to an input image to identify the characteristics of the image and checks the quality indicators for the identified characteristics from the quality indicators for various image characteristics that are pre-set. Then, for each image block constituting the image, the aforementioned methods 1 through 3 and method 4 are applied individually to each, and the maximum approximation strength is determined such that the quality does not fall below the identified quality indicator.

[0146] And, when the approximation strength is determined by each method determined in this way, the highest approximation strength among the determined approximation strengths is selected as the approximation strength for the corresponding image block. As an example of a specific numerical value, for any image block, if the maximum approximation bit satisfying the quality index confirmed for the corresponding image characteristic is determined to be 3, 4, 5, and 3 in the above methods 1 through 4, respectively, then the approximation unit (101) applies an approximation strength of 5 to the corresponding image block.

[0147] In one embodiment according to the present invention, selective approximation of pixel data may be based on correlation between temporally consecutive images in addition to approximation based on correlation between pixels within the same image as described above (this is referred to as 'spatial approximation' to distinguish it from the approximation method below). FIG. 6 schematically illustrates an example of this process (referred to as 'temporal approximation' in this specification).

[0148] For temporal approximation, the approximation unit (101) is at the current time (t k While selectively performing spatial approximation on the pixel data of an image input to ) in the manner described above, the pixel data to which the selectively spatially approximated is applied are pre-specified image blocks of a certain size (pb t[k]Whenever a ) is formed, for each pixel data belonging to that image block, the previous time point (t) previously stored in the frame buffer (102) k-1 Image block (pb) corresponding to the position on the image of ) t[k-1] The difference in pixel values ​​with the pixel data of ) is calculated for each (S61).

[0149] Current image block (pb t[k] For pixel data belonging to ), the previous image block (pb t[k-1] When the difference in pixel values ​​with ) is individually calculated, the approximation unit (101) compares each of the pixel value differences with a preset first limit value (S62), and when all of the pixel value differences are less than or equal to the first limit value, the approximation unit (101) [compares] the image block (pb t[k] Regarding ), the outer portion is limited, and the difference in each pixel value of the pixel data belonging to the outer portion is compared with a pre-specified second limit value (S63). Here, the outer portion may be only the pixels directly adjacent to the boundary of the image block, or N from the rectangular boundary B ( N B ... may be pixels belonging to a range that extends inward into the block up to, for example, 2 or 3 pixels. The outer part used in this sense is called the 'boundary zone'.

[0150] In the above comparison (S63), if it is confirmed that all pixel value differences of the boundary pixel data are less than or equal to the second limit value (S64), then the current image block (pb t[k] The pixel data of ) is not written to the frame buffer (102). Instead, the corresponding image block (pb t[k] For ), the previous image block (pb t[k-1] Only information indicating replacement with ) (hereinafter, this information is referred to as the 'temporal approximation flag') is recorded (S65). That is, the current image block (pb t[k]Regarding the pixel data of ), in the subsequent visual representation process, the previous image block (pb t[k-1] It is expressed by replacing it with the pixel data of ) (this process will be described in detail later). This means that the pixel data of the current image block is approximated with the pixel data of the previous image block. In other words, it means that temporal approximation has been applied to the current image block.

[0151] In the above description, it is preferable that the second limit value applied to the boundary zone of the image block has a smaller value than the first limit value applied to the entire image block. That is, when determining whether to apply temporal approximation to the image block, it is preferable to apply a stricter standard to the difference in pixel values ​​of the boundary zone to the entire image block.

[0152] The above temporal approximation flag may be 1 bit of information added separately for each image block (this information may have a value of 1 indicating On if the current image block is omitted from writing to approximate the previous image block, for example, and a value of 0 indicating Off otherwise).

[0153] The approximation unit (101) is such that if the difference in pixel value between any of the pixel data of the current image block and the pixel data of the previous image block is greater than the first limit value, or if the difference in pixel value between any of the pixel data of the boundary zone is greater than the second limit value, then the corresponding image block (pb t[k] Regarding ), temporal approximation is not applied and it is written to the above frame buffer (102).

[0154] In another embodiment according to the present invention, instead of applying temporal approximation when the difference in pixel values ​​for pixel data within the current image block is all less than or equal to the first limit value and at the same time the difference in pixel values ​​for each of the pixel data in the boundary zone is all less than or equal to the second limit value, temporal approximation may also be applied to the current image block when either of the two conditions is satisfied, that is, when the difference in pixel values ​​between the pixel data of the current image block and the previous image block is all less than or equal to the first limit value (S62), or when the difference in pixel values ​​for each of the pixel data in the boundary zone is all less than or equal to the second limit value (S64).

[0155] In another embodiment according to the present invention, the current image block (pb t[k] The image block immediately preceding ) (pb t[k-1] Temporal approximation can be selectively applied based on the total pixel value difference (or the difference in the average pixel value) with respect to ). In this embodiment, the pixel value difference between the pixel data belonging to the current image block and the previous image block (the sum of all individual pixel value differences or the sum divided by the number of pixel data in the image block) is calculated, and the calculated difference is compared with a preset third limit value; if it is less than or equal to the third limit value, the current image block (pb t[k] Apply temporal approximation to ).

[0156] In one embodiment according to the present invention, the first to third limit values ​​presented for the selective application of temporal approximation may also be used for the selective application of spatial approximation as described above. That is, for any image block, if the block difference component obtained from the difference of at least some pixel values ​​with the previous image block (the difference in pixel values ​​between each corresponding pixel for comparison with the first limit value, the sum or average of the differences in boundary pixel values ​​for comparison with the second limit value, or the sum or average of the differences in all pixel values ​​for comparison with the third limit value) is less than or equal to the target limit value to be compared, spatial approximation is applied to the image block; otherwise, it is stored in the frame buffer (102) without approximation. Of course, regarding the first to third limit values, the magnitude of the values ​​applied for temporal approximation and spatial approximation may be different.

[0157] The priority of application for the first to third limit values ​​may vary depending on the conditions set in the display device (100). For example, if a mode prioritizing image quality is set, the first limit value is set to be applied before or primarily over the second and third limit values, and if a mode prioritizing resource saving over image quality is set, the priority of application is set to be the opposite. If the limit value is set relatively higher, the frequency of application increases, so the effect of increasing the priority of application can be obtained. Of course, the priority of application may also be directly determined according to the conditions set.

[0158] The priority of application of the first to third limit values ​​mentioned above can naturally be applied even when these limit values ​​are used in the aforementioned temporal approximation.

[0159] In addition, in one embodiment according to the present invention, the first to third limit values ​​may be dynamically varied. As factors acting on the dynamic variation of the limit values, the motion vector of an image frame or image block applied to the display device (100), or the frame rate of the image, may be adopted. As a specific example, if the variability of the image increases, for example, if the motion vector increases or the frame rate increases, at least one of the first to third limit values ​​may be changed to a larger value. When the limit value is changed to a higher value, the frequency of approximation for pixel data increases compared to before the change.

[0160] In one embodiment according to the present invention, the temporal approximation described above may also be applied on a frame basis. In other words, an image block of arbitrary size described above may correspond to a single frame. In this case, if the condition (or conditions) regarding the difference in pixel values ​​between the current frame and the previous frame described above is satisfied, the current frame is not recorded in the frame buffer (102), but instead only a temporal approximation flag indicating replacement with the previous frame is recorded. In this embodiment, whether temporal approximation is applicable to the current frame is determined as described above, and if it is determined to be applicable, temporal approximation is applied to the current frame; if it is determined not to be applicable, the current frame is divided into image blocks of a specified size, and temporal approximation is selectively applied to the divided image blocks as described above.

[0161] In one embodiment according to the present invention, the temporal approximation operation described above may be stopped at a certain period, for example, every certain number of image frames. More specifically, the approximation unit (101) counts the input image frames, and whenever the counted value becomes a multiple of, for example, 6 or 30, it does not check whether temporal approximation is applied to the blocks (or the frame) belonging to that frame. That is, it performs only the spatial approximation described above on the pixel data within that sequence of frames. Accordingly, all pixel data within that sequence of frames (some may be approximated pixel data) are transmitted to the display control unit (110) through the frame buffer (102).

[0162] Through the periodic interruption of such temporal approximation, all pixel units of the panel (120) are refreshed. That is, each pixel unit is charged with the pixel values ​​of each pixel data from a newly input image frame, without the process of restoring the previous pixel value from the current residual voltage of any pixel unit, which will be explained in detail later, and recharging it with the restored value. Of course, even at this time, for each approximated pixel data, the value of the discarded LSB bit estimated based on the residual voltage of the corresponding pixel unit may be supplemented to charge the corresponding pixel unit. This operation will be explained later.

[0163] In one embodiment according to the present invention, the determination of whether to apply the aforementioned spatial approximation and temporal approximation to pixel data within an image block of arbitrary size and / or the approximation strength may be performed by a neural network trained by a specific deep learning algorithm. To this end, a neural network trained by a deep learning algorithm is provided and executed by the approximation unit (101).

[0164] The neural network provided in the approximation unit (101) above, for example, a deep convolution neural network, may be composed of a module that performs spatial approximation by multi-layeredly connecting approximation coefficients (specific example, sets of kernel coefficients) applied to individual pixel data or regions, and a module that performs spatial approximation by multi-layeredly connecting approximation coefficients applied to the entire image or divided image blocks, and the approximation coefficients forming the inter-layer connections of each module are values ​​set by learning so that the image quality resulting from the approximation of numerous sample images is adjusted to the user's preference.

[0165] Meanwhile, while the approximation unit (101) selectively performs spatial approximation and temporal approximation (in this specification, the term 'approximation' is used to include both spatial approximation and temporal approximation unless it is a description of an embodiment in which it cannot be applied) on the pixel data of the input images as described above and records it in the frame buffer (102), the data supply unit (103) transmits the pixel data of the images recorded in the frame buffer (102) sequentially, for example, according to the order of the raster scan method and according to the feeding timing determined for screen composition, to the display control unit (110).

[0166] The data supply unit (103) at the time, in the process of transmitting pixel data, if the pixel data is approximated, refers to the corresponding approximation strength information and pads 0 (or 1) equal to the number of bits of the discarded LSB to form pixel data of a complete number of bits and transmits it.

[0167] In another embodiment according to the present invention, instead of padding of 0 (or 1), approximate intensity information may be transmitted to the display control unit (110). In this embodiment, approximate intensity information corresponding to a series of pixel data corresponding to a scan line may be transmitted in the order of the pixel data, for example, within a porch period based on a horizontal or vertical synchronization signal between the time when the transmission of pixel data of the previous scan line is completed and the time when the transmission of pixel data of the current scan line is to begin (or within a time interval between the time when the transmission of pixel data of the current scan line is completed and the time when the transmission of pixel data of the next scan line is to begin). Alternatively, it may be transmitted during a blank frame period based on a vertical synchronization signal preceding or immediately after the image frame containing the corresponding pixel data. In this specification, the time of the porch period or blank frame period preceding the time when pixel data is to be transmitted in the supply timing is referred to as an 'idle interval'. Of course, in an embodiment where a separate signal line is provided between the data supply unit (103) and the display control unit (110) in addition to the pixel data transmission line for transmitting approximate strength information, the approximate strength information is transmitted to the display control unit (110) at the same time as the pixel data transmission.

[0168] If approximate strength information is transmitted based on the synchronization signal by being carried on a horizontal or vertical synchronization signal or through a separate signal line, it is not necessary to add identification information indicating the image block, etc., to which the approximate strength of the information should be applied. This is because the display control unit (110) can distinguish which image block, i.e. which pixel data, the approximate strength information to be applied is determined by sequentially mapping the approximate strength information transmitted on the synchronization signal or through a separate signal line to the series of pixel data transmitted from the data supply unit (103) based on the horizontal or vertical synchronization signal. Therefore, the overhead for transmitting approximate strength information to the display panel side can be minimized.

[0169] If temporal approximation is applied according to the aforementioned embodiment, the data supply unit (103) first checks the temporal approximation flag for the pixel data in the order to be transmitted at the current supply timing, that is, for the image block to which the pixel data belongs, and if the flag is On, does not transmit the corresponding pixel data. Alternatively, meaningless pixel data may be transmitted at that timing. Then, during the valid interval preceding the time of transmission of the corresponding pixel data, a temporal approximation flag indicating that pixel data is missing due to temporal approximation is transmitted in advance to the display control unit (110).

[0170] The display control unit (110) applies pixel data, which is sequentially transmitted from the data supply unit (103) in accordance with the supply timing as described above, to the pixel units provided in the display panel (120) in units of scan lines. More specifically, when pixel data corresponding to one scan line is input, the display control unit (110) applies a voltage corresponding to each data to each source line and drives the gate line corresponding to the scan line to be rendered on the current screen, thereby causing the pixel values ​​of the series of pixel data corresponding to one scan line to be transmitted to the cell capacitors (C) of the pixel units. C It is charged to the cell capacitor. In the display panel (120), a color (or gradation) according to the voltage of the cell capacitor charged in this way is displayed.

[0171] According to one embodiment of the present invention, when approximate intensity information for approximated pixel data is provided from the data supply unit (103), the display control unit (110) applies the value to the pixel units belonging to the scan line to be rendered at the current time to the approximated pixel data, that is, to the pixel data for which the corresponding approximate intensity information is not zero and is received together with the beginning of the pixel data or received separately before through an idle interval, restores the number of LSB bits indicated by the approximate intensity information for the pixel data and adds them to the approximated pixel data to preserve the approximated pixel data, or preserves all bits as previous bits according to the approximate intensity information indicating complete approximation (all bits omitted by approximation), and then applies the pixel value of the preserved pixel data to the corresponding pixel unit.

[0172] FIGS. 7a and 7b schematically illustrate a configuration and process according to the present embodiment in which bits that are selectively approximated for pixel data, namely discarded LSBs or all bits, are preserved and applied to the corresponding pixel unit, and this process will be explained in detail below.

[0173] FIG. 7a illustrates a data processing module within the display control unit (110) that constitutes pixel data to be charged to the pixel unit, that is, to be rendered on the screen. This processing module is composed of a pixel data preservation unit (111), a rendering data selection unit (112), and a data driving unit (113). Furthermore, this processing module may be individually provided for each pixel unit belonging to a row of the panel (120). That is, a number of processing modules equal to the number of source lines, SNum, may be provided in parallel.

[0174] The pixel data preservation unit (111) performs an operation to preserve lost information of the pixel data using the residual voltage or residual charge amount currently remaining in the corresponding pixel unit to be rendered (hereinafter, the residual voltage or residual charge amount of the pixel unit is also referred to as the 'residual value') based on the applied approximation information. The process of preserving this lost information will be explained in detail later. Furthermore, the approximation information refers collectively to the approximation strength information and the temporal approximation flag mentioned earlier, and in an embodiment where approximation strength information is not provided for pixel data to which spatial approximation is applied, it naturally refers only to the temporal approximation flag.

[0175] The rendering data selection unit (112) is configured to select and output one of the applied pixel data and the pixel data in which information lost by the data preservation unit (111) is preserved (hereinafter referred to as 'substitute data') based on the applied data selection signal (dsel_ctl), and the data driving unit (113) drives the source line to which the corresponding pixel unit is connected with an analog voltage corresponding to the selected output pixel data.

[0176] The above display control unit (110) checks whether there is any approximation in the pixel data to be applied to the pixel units belonging to the scan line to be applied to the panel (120) at the current time. That is, it checks whether there is pixel data in which the LSB is partially discarded (or filled with a specific bit value) by spatial approximation in the pixel data to be applied, or a part in which the corresponding pixel data is missing by temporal approximation (i.e., a part in which the entire part is discarded and only a temporal approximation flag is applied). In this specification, pixel data that is partially missing because a part of the original LSB is discarded by spatial approximation and a completely missing empty part in which the original pixel data is entirely missing by temporal approximation and only the fact of the missing part is known as information such as a flag are collectively referred to as ‘missing data’.

[0177] If there is missing data in which the original information regarding color or gradation is wholly or partially lost due to approximation among the pixel data to be applied to the current scan line, the display control unit (110) sets the data selection signal (dsel_ctl) to a value that selects and outputs the data output from the pixel data preservation unit (111) for the rendering data selection unit (112) of FIG. 7a assigned to the pixel column where the missing data is to be recorded. Of course, if there is no missing data, the data selection signal (dsel_ctl) is naturally maintained at a value that allows the applied pixel data to be selected and output as is for the corresponding processing module. Of course, as described above, in an embodiment where the data supply unit (103) does not provide approximation strength information for pixel data in which some LSB bits are missing, the data selection signal (dsel_ctl) is naturally maintained at a value that allows the applied pixel data to be selected and output as is for the missing data as well.

[0178] Along with the selection of the output of the rendering data selection unit (112), the display control unit (110) applies each pixel data to the pixel data preservation unit (111) and the rendering data selection unit (112) of the processing module of FIG. 7a in sequence, and also applies approximation information for each pixel data to the pixel data preservation unit (111).

[0179] After applying pixel data and approximation information for each processing module of FIG. 7a and setting a selection signal (dsel_ctl), the display control unit (110) detects the charging voltage remaining in the pixel units belonging to the scan line on the panel (120) to be rendered, that is, the row of pixel units (hereinafter referred to as the 'target row,' and for a specific example, it is assumed that this target row is in the m (1≤m≤M)th order), that is, the charging voltage remaining in each cell capacitor of the pixel units, and applies it to the pixel data preservation unit (111) of each processing module of FIG. 7a.

[0180] To this end, first, the display control unit (110) determines which line segment a row position m belongs to in a source line. This is determined from setting information regarding how many pixel units each line segment is divided into. When the line segment thus identified is confirmed to be the k-th, the display control unit (110) [determines] the upper and lower switches (SW) of the k-th line segment in each source line. i_k-1 and SW i_k Both switching lines (scL) connected to , i=1,2,..,SNum) i_S[k-1] , scL i_S[k] Signals are applied to the (i=1,2,..,SNum) (naturally, only one switch when the corresponding line segment is at the top or bottom) to turn off the corresponding switches respectively (31 in FIG. 3).

[0181] Each source line (S i After turning off both switches for , i=1,2,쪋,SNum), the gate line (G) corresponding to row position m. m ) driving the cell transistors (Tr) of all pixel units (32) belonging to the target row C) are each short-circuited. Accordingly, the voltage remaining in the cell capacitors of all pixel units belonging to the target row, as well as the pixel unit (32) shown in FIG. 3, is transferred to the recovery lines (rL) connected to each line segment. i_S[k] Each is discharged along i=1,..,n,..,SNum). When applying charge sharing to detect the remaining voltage (this is described in detail below), a voltage of a certain magnitude may be applied to the corresponding recovery line, and the voltage applied may be greater than the residual voltage currently remaining in the pixel unit. In this case, the residual voltage is not discharged through the corresponding recovery line but is rather charged. Therefore, the phrase 'discharge through the recovery line' below should be interpreted as charging through the corresponding line depending on the applicable embodiment.

[0182] In one embodiment according to the present invention, in the case of a specific type of display panel, for example, an OLED or micro-OLED type panel, a separate transistor element is used to discharge the voltage currently remaining in the pixel unit and check the residual value, and a cell capacitor (C C It may also be circuitously connected to ), and in this embodiment, the gate line (G) is a control signal line for selectively opening and closing the transistor element. i It is provided in parallel with , i=1,..,M).

[0183] The above display control unit (110) is the recovery lines (rL i_S[k]The magnitude of the voltage discharged along each of the lines (i=1,..,n,..,SNum) is detected sequentially or in parallel, and the pixel value immediately applied to each pixel unit of the target row is estimated from the detected residual value. The voltage discharged along each of the recovery lines, i.e., the voltage remaining in the cell capacitor, is detected by the following method by applying the principle of charge sharing.

[0184] FIG. 8a shows the voltage (rmV) remaining in the pixel unit located at a specific column (the nth column in order) of a target row, according to one embodiment of the present invention. n,m ) the corresponding recovery line (rL n_S[k] It shows the detection method through ).

[0185] The above display control unit (110) is the corresponding gate line (G m After the operation of ), each recovery line (rL i_S[k] , i=1,..,n,..,SNum) connected to a detection capacitor (C Det The voltage (V) charged to ) Det Detects the gate line (G m When ) is driven, the amount of charge (Q) accumulated in each pixel unit of the target row n,m(O) The voltage is redistributed according to the capacitance of the capacitor, and the voltage becomes balanced.

[0186] That is, Q n,m(O) = Q Det + Q n,m It becomes.

[0187] Therefore, according to the above equation, the voltage remaining in the corresponding pixel unit (rmV n,m ) is a detection capacitor (C) that is in equilibrium after charge redistribution. Det ) voltage (V Det If ) is detected, it can be obtained according to the following equation [1] (801).

[0188] rmV n,m = (1+C Det / CC )·V Det Equation [1]

[0189] Equation [1] is a detection capacitor (C Det This is about detecting the residual voltage of the corresponding pixel unit after charge redistribution when the voltage is not charged in ).

[0190] In another embodiment according to the present invention, the display control unit (110) detects a predetermined initial voltage using the detection capacitor (C Det While charged in ), the corresponding gate line (G m After driving the ) to redistribute the charge, the detection capacitor (C) that has reached equilibrium is Det The residual voltage of the corresponding pixel unit can also be detected by detecting the voltage of the ). This can be applied when you want to increase the residual voltage to a voltage within an appropriate range that is easy to detect. In this case, the residual voltage is calculated by applying the following equation [2].

[0191] rmV n,m + C Det ·V Det(0) / C C = (1+C Det / C C )·V Det Equation [2]

[0192] Here, V Det(0) is a detection capacitor (C) before charge redistribution for detecting residual voltage. Det It is the initial voltage that charges the ).

[0193] However, when detecting the voltage remaining in the corresponding pixel unit according to Equation [1] or [2] and detecting the voltage charged to the corresponding pixel unit, i.e., the applied pixel value, at least two things must be considered and applied to the detection.

[0194] First, the influence of parasitic capacitance caused by pixel units commonly connected to the line segment and parasitic resistance of the recovery line is considered.

[0195] FIG. 8b shows the voltage (rmV) remaining in the pixel unit located at a specific column (the nth column in order) of the target row. n,m ) is the corresponding recovery line (rL n_S[k] ) through the detection capacitor (C Det ) and when charge is redistributed, parasitic capacitance (C par_T This represents an equivalent circuit showing the effect of ) on the charge redistribution. As illustrated, the residual value (rmV) currently remaining in the pixel unit n,m The amount of charge due to ) is detected by the detection capacitor (C Det In addition to ), parasitic capacitance (C par_T Since it is distributed to ) (82), this must be reflected in the residual value obtained by Equation [1] or [2].

[0196] Recovery line (rL) n_S[k] Parasitic capacitance (C) caused by commonly connected pixel units acting on ) par_T The magnitude of the charge accumulated by ) is C par_T ·V Det / C C Therefore, the above display control unit (110) is a detection capacitor (C Det Voltage detected at ) (V Det Regarding ), parasitic capacitance (C) for cell capacitance par_T The ratio α of ) CT (=C par_T / C C Correction is performed by adding ) to the residual voltage obtained by Equation [1] or [2]. Of course, this correction is the above ratio α CT It may also be accomplished by an adder circuit having circuit characteristics that reflect it.

[0197] Parasitic capacitance (C) generated by pixel units commonly connected to a single line segment par_T If ) is too large, the detected residual voltage (V Det) becomes that much smaller. This can lead to inaccuracy in detecting residual values. Therefore, it is desirable to divide a single source line into multiple line segments by considering the relative size of the parasitic capacitance by each line segment to the cell capacitance.

[0198] Of course, if the parasitic capacitance caused by all pixel units commonly connected to a single source line is of an acceptable size compared to the cell capacitance, there is no need to configure the source line in a form where switches are inserted in series as in the above-described embodiment. In other words, there is no need to segment the source line.

[0199] And, as exemplified in FIG. 3, an arbitrary source line (S n ) is divided into multiple line segments, and a recovery line (rL) directly connected to each line segment is n_S[i] When each is equipped with ), since the length of each recovery line to the point where the circuit part for detecting residual value is located, for example, to the outer edge of the panel (120), is different, the size of the parasitic capacitance component according to the length of the recovery line is bound to be different, and this parasitic capacitance component may have a size that cannot be ignored compared to the parasitic capacitance component caused by the pixel units commonly connected to the line segment.

[0200] Accordingly, the display control unit (110) has a ratio α to the capacitance for the commonly connected pixel units. CT The ratio of capacitance α according to the length of the corresponding recovery line CL ( =C par_L / C C , C par_L is the ratio α including the parasitic capacitance caused by the corresponding recovery line. C ( =α CT + α CLBy providing circuit or operation information that is reflected differently for each line segment, the circuit or operation information set for the recovery line connected to the line segment to which the pixel units of the corresponding scan line belong can be selected and used.

[0201] Since the parasitic capacitance component cannot be ignored relative to the cell capacitance, the ratio α of parasitic capacitance to cell capacitance CT Where circuit or computational information is provided to reflect, or parasitic capacitance ratio α C Circuit or operation information to reflect is retrieved line (rL n_S[i] In the case where provided for each of the ), a detection capacitor (C) as exemplified in FIG. 8a. Det Without ) and as shown in FIG. 8c, parasitic capacitance (C) acting on the recovery line par = C par_T or C par_T +C par_L ) can be used for detection through charge redistribution.

[0202] That is, when using the parasitic capacitance acting on the recovery line as exemplified in FIG. 8c without connecting a separate capacitor to the recovery line, α, which is pre-reflected in the circuit or provided as computational information as previously explained, CT or α C Using the following equations [3] and [4], the residual voltage, i.e., the residual pixel value, is obtained from the voltage that has been balanced.

[0203] rmV n,m = (1+α CT )·V Det , or

[0204] rmV n,m + α CT ·V Det(0) = (1+α CT )·V Det Equation [3]

[0205] rmV n,m = (1+αC )·V Det , or

[0206] rmV n,m + α C ·V Det(0) = (1+α C )·V Det Equation [4]

[0207] Secondly, the degree to which the voltage charged in the cell capacitance of the pixel unit leaks over time is considered. The degree of leakage over time is the cell capacitor (C C The time constant T, which is the product of ) and the parasitic resistance forming a closed circuit. RC It is determined by. The voltage of the applied pixel value changes as time progresses due to β leak =exp(-t / T RC Since leakage occurs at a ratio, the residual value (rmV) detected according to the method described above n,m The residual ratio (1-β) obtained by substituting the time according to the frame rate (e.g., 1 / 60, 1 / 120 second, etc.) into ). leak0 , β leak0 =exp(-1 / (60·T RC )), or exp(-1 / (120·T RC By multiplying by the reciprocal of (or amplifying) etc., one can obtain the pixel value estimated to have been applied in the previous frame. Of course, the above residual ratio (1-β leak0 )(or its reciprocal) is set in the display control unit (110) as circuit or operation information.

[0208] Accordingly, the above display control unit (110) is, the corresponding gate line (G m After driving ), each recovery line (rL i_S[k] Each detection capacitor (C) is connected through , i=1,..,n,..,SNum). Det ) or parasitic capacitance (C par_T or C par The voltage charged in ) (V in the aforementioned description) Det(0)Depending on the size, it may be a discharged voltage. By applying the operation according to the two items described above to (or adding, amplifying, etc.), the pixel values ​​that are presumed to have been applied to the pixel units of the target row immediately before are each obtained and applied to the pixel data preservation unit (111) of the processing module of FIG. 7a that corresponds in order.

[0209] In the case where an embodiment is applied to spatially approximated missing data that preserves the value of the discarded LSB bit in the data, the pixel data preservation unit (111) of each processing module determines whether to preserve the currently applied missing data based on the applied approximation information when the estimated previous pixel value (71) is input, as shown in FIG. 7b. When the authorized approximation information indicates non-zero approximation strength information (i.e., indicates that the corresponding pixel data has been approximated), the pixel data preservation unit (111) refers to the approximation information and supplements the LSB discarded by approximation (or the value lost as padding of 0 or 1) with the corresponding LSB part (71a) in the input estimated pixel value (71) (or replaces it with the value of that part (71a)), thereby constructing substitute data (72) that preserves the part of information lost by approximation in the authorized defective data and outputs it to the rendering data selection unit (112) (p70).

[0210] In one embodiment according to the present invention, when the pixel data preservation unit (111) preserves the portion of information lost due to approximation, it may selectively do so by comparing the portion of the missing data that is not approximated with the previous pixel value estimated as described above. For example, regarding missing data in which 5 bits are discarded by spatial approximation, when the remaining upper 3 bits and the upper 3 bits of the previous pixel value estimated from the residual voltage of the corresponding pixel unit are the same value, substitute data constructed by supplementing (or replacing bits filled with 0 or 1) the 5 bits discarded as described above with the LSB 5 bits of the estimated pixel value is output, and when the upper 3 bits are different values, the missing data is output as is to the rendering data selection unit (112) without preserving the loss information. At this time, when the missing data has only upper bits, pixel data filled with 0 or 1 for the missing lower bits is output.

[0211] The preservation of selective loss information through comparison of the upper bits as described above is intended to prevent the possibility of preservation that deviates significantly from the original pixel value when the pixel value approaches the median value. For example, if the pixel value applied immediately before is "0111 1111" and the pixel value to be applied now was originally "1000 0001" but is approximated with an approximation strength of 7 so that only the most significant bit of 1 is transmitted, then when the residual value is detected and preserved, substitute data of "1111 1111" may be constructed, and in such a case, it may deviate significantly from the original pixel value.

[0212] Meanwhile, if the authorized approximation information indicates an approximation strength of 0, the pixel data preservation unit (111) does not perform any operation on the authorized pixel data.

[0213] If the authorized approximation information indicates that the corresponding pixel data is missing due to temporal approximation (i.e., not received as valid pixel data from the data supply unit (103)), i.e., that the original information is completely lost, the pixel data preservation unit (111) constructs substitute data that preserves the lost information using all of the previous pixel values ​​(71) obtained by estimating based on the residual value detected in the corresponding pixel unit as described above, and outputs it to the rendering data selection unit (112).

[0214] Meanwhile, a data selection signal (dsel_ctl) that is set as a value according to whether the pixel data is approximated is applied to the rendering data selection unit (112) by the display control unit (110). In the case of missing data to which approximation is applied, the pixel data preservation unit (110) selects substitute data that preserves the lost information and applies it to the data driving unit (113), and in the case of non-applied data, the applied pixel data is selected and applied to the data driving unit (113).

[0215] As described above, the display control unit (110), with respect to pixel data to be applied to a series of pixel units corresponding to the current scan line (i.e., corresponding to row position m), if the data is approximated missing data, substitute data is applied to the data driving unit (113) as is otherwise, and if the data is not, the input pixel data is applied as is to each data driving unit (113), then activates an enable signal (drv_En) to each data driving unit (113) so that the voltage corresponding to the applied data is applied to each source line (S i After making it output to , i=1,..,n,..), the corresponding gate line (G m By driving ), it is made to be visually displayed in the target row on the panel (120).

[0216] In another embodiment according to the present invention, missing parts may not be preserved for approximated missing data. Instead, the display control unit (110) does not drive the pixel unit to which the bits of the missing parts are to be expressed, for example, the source of the unit. If all the pixel data of any row is missing data, the gate driving for the pixel units of that row is not performed. This embodiment has the advantage of reducing power consumption by utilizing the fact that missing bits do not significantly affect image quality even if they are not expressed as color (or gradation).

[0217] In the above-described embodiment, recovery lines (rL) which are circuit lines for detecting the residual value of an arbitrary pixel unit i_S[k] , i=1,..,n,..,SNum) is a single source line (S i For i=1, 2, ..., n, ...), they are arranged in parallel with each other for each line segment and are provided in a form that extends individually to the point where the circuit part for detection is located, for example, to the outer edge of the panel (120). In this embodiment, since recovery lines must be provided for each source line as many times as the number of line segments, the area occupied by the recovery lines becomes larger. This becomes a disadvantage for display devices using backlights.

[0218] For this reason, in another embodiment according to the present invention, the recovery line may be provided hierarchically to reduce the occupied area by the recovery line. FIG. 9 is an arbitrary source line (S) according to the present embodiment. n Regarding ), at the point where there is a circuit section for detecting residual values, a single main circuit line link (rL n_0 Circuit line links (rL) starting at the top n_i_j , i=1,..,L, j=1,2,..) branch hierarchically, and the lowest circuit line links (rL n_L_jIt shows that a recovery line is provided in a tree structure form, with each of the line segments (, j=1,2,..) connected.

[0219] And, each branching point is equipped with a selection switch (40) for selecting a path, and this selection switch is a selection signal (SbCtl) applied to a selection terminal (ST). i Depending on the value of , i=1,2,..,L), both ends (CT U , CT D One of them is electrically connected to the MT.

[0220] The above selection signal (SbCtl i , i=1,2,..,L) is the number of layers of circuit line links constituting the return line (more precisely, when LN is the number of line segments, 2 x Each source line is provided with the smallest integer of x satisfying the inequality >=LN, and one selection signal (SbCtl K , K is any value from 1 to L) are selection switches (Sb) corresponding to the same level in the hierarchy. K_i It is electrically connected so that it is commonly applied to , i=1,2,..).

[0221] According to such a circuit connection structure, the series of selection signals (SbCtl i When the values ​​of , i=1,2,..,L) are respectively assigned, both ends (CT) of the selection switches for each level are assigned accordingly. U ,CT D One of ) is selected and connected to the parent (MT), and the above parent circuit line link (rL n_0 An effective electrical path is formed starting from ) to any one line segment, and the electrical path thus formed forms a recovery line for that line segment.

[0222] Accordingly, the display control unit (110) identifies which line segment a pixel unit (hereinafter referred to as the 'target unit') belongs to, which pixel data needs to be rendered at the current time and therefore needs to detect a residual value at the current time, and appropriately selects the values ​​of the selection signals accordingly, thereby [determining] the main circuit line link (rL n_0 After forming an effective electrical path from ) to that line segment, through the path of the circuit line links thus formed, i.e., the recovery line, the detection capacitor (C Det The residual value is obtained by detecting the voltage resulting from the redistribution of charge in ).

[0223] As illustrated in FIG. 9, if the recovery line is configured as a hierarchical structure for each source line, the area occupied by the recovery line per source line is reduced compared to the previously described embodiment. For example, when one source line is divided into 64 line segments, in the embodiment according to FIG. 3, 64 recovery lines, i.e., 64 circuit lines, occupy the area, but in the embodiment according to FIG. 9, the area is occupied by 6 layers of circuit line links and 6 selection signals for the recovery line, so the area occupied by the circuit lines is significantly reduced.

[0224] In one embodiment according to the present invention, any source line (S n A switch (SW) for line segmenting ) n_i Switching lines (scL) to which a signal to turn On / Off , i=1,2,쪋) is transmitted n_S[i] For , i=1,2,3,..), it can also be constructed in the same way as the hierarchical structure of the recovery line according to the example of Fig. 9.

[0225] Up until now, a source line (S) for applying pixel values ​​to pixel units arranged in any column on the panel (120) iFor , i=1,..,n,..,SNum), as exemplified in FIG. 3, a plurality of switches (SW i_j The source line is segmented by inserting ) and in each line segment, a recovery line (rL) for detecting the residual value of the pixel unit is included. i_S[j] ) is connected, and the source line (S i Switching lines for transmitting signals to control the opening and closing of switches inserted in ) (scL i_S[j] The principle and technical concept of the present invention for selectively reusing residual pixel values ​​of a pixel unit based on an embodiment having a circuit configuration in which each of the switches is connected has been specifically explained. However, in another embodiment according to the present invention, the pixel unit layer of the panel (120) is, as illustrated in FIG. 10, each pixel column (lsS i , i=..,n,..) are line segments (ls) that are previously electrically disconnected from each other (1010). n_i It is composed of , i=..k-1,k,k+1,..) and each line segment may be provided with a plurality of pixel units (pxG) commonly connected as described above.

[0226] In the embodiment according to FIG. 10, the line segments forming a single column are disconnected from each other, so they cannot be used to apply a voltage corresponding to pixel data to a pixel unit. Instead, the corresponding pixel column (lsS n Line segments belonging to ) (ls n_i , signal line(wrL) connected to i=..k-1,k,k+1,..) n_S[i], i=..,k,k+1,..) is used for the purpose of reading the residual value of the pixel unit, in the same manner as the use of the recovery line in the aforementioned embodiment, and is also used for the purpose of applying a voltage corresponding to the pixel data to the pixel unit. In this embodiment, the display control unit (110) determines which line segment the scan line of the pixel unit to which the pixel value of the current pixel data is to be applied belongs to, and the signal line (wrL) corresponding to that order n_S[k] A circuit that applies pixel values ​​through ) is provided.

[0227] And, the process of reading the residual value of any pixel unit is the same as in the previously described embodiment. However, in the embodiment according to FIG. 10, since the pixel units arranged in each pixel column are divided into multiple groups and electrically disconnected in advance, the operation of selectively opening and closing the switch is not involved.

[0228] Also, electrically disconnected line segments (ls n_i , i=..k-1,k,k+1,..) to the signal line (wrL) connected for the application of pixel values ​​and detection of residual values n_S[i] , i=..,k,k+1,..) may be configured with signal line links and path selection switches in a hierarchical tree structure as exemplified in FIG. 9.

[0229] Meanwhile, in the method described above for detecting the residual value of an arbitrary pixel unit to obtain the pixel value applied immediately prior, the parasitic capacitance considered, as well as the cell capacitance of the pixel unit, fluctuates in value due to the influence of operating environments such as temperature. Therefore, the capacitance ratio α described above C or α CT If the previous pixel value is obtained from the residual value by applying a circuit element or calculation method fixed to a specific value, the accuracy of the obtained pixel value may be degraded as the operating environment, such as display panel process parameters and operating power, fluctuates.

[0230] Accordingly, in another embodiment according to the present invention, in order to maintain the accuracy of the pixel value estimation applied immediately prior to the variability of the operating environment, the panel (120) is provided with a set (200) of reference pixel units (hereinafter abbreviated as 'reference unit set') as exemplified in FIG. 11a, formed together with pixel units that express color (or gradation).

[0231] The above reference unit set (200) is exposed to the same environment as the operating environment of the pixel units and the circuits attached thereto, so that the change in electrical characteristics due to the change in the operating environment of the panel (120) appears the same as that of the pixel units, etc.

[0232] Looking specifically at the configuration of the reference unit set (200) according to one embodiment of the present invention, the reference unit set (200) comprises reference lines (rfS) of a number that is one less than the number of values ​​that pixel data can have (256 in the case of pixel data with a bit width of 8) (in the case of an embodiment in which a predetermined voltage is applied to the pixel unit even for a pixel value of 0, it may be a number equivalent to the number of values ​​that pixel data can have). i Each of , i=1,..,255) has a structure in which reference units (refU) are commonly connected as many times as the number of source lines (SNum) for applying pixel values ​​to the pixel units. That is, the reference unit set (200) is composed of, for example, 255 × SNum reference units. Here, each reference unit (refU) has a circuit structure exactly identical to that of a pixel unit (pxU) and can be formed together when the pixel unit is formed on a semiconductor chip.

[0233] And, a series of reference units, for example, 255, connected to different reference lines, are a single selection line (rfG i The gates are commonly connected to , i=1,..,SNum).

[0234] Also, each reference line (rfS i In the ), one or more switches (rfSW), such as transistors, are inserted in series, and the switches are turned On or Off according to a signal applied to the switches (rfSW) by the display control unit (110). When turned Off, the reference line is partially segmented. This is to reduce the influence of parasitic capacitance caused by other commonly connected reference units, as described regarding the segmentation of the source line. The switches (rfSW) are formed by being inserted into the reference line such that the number of reference units commonly connected to each reference line segment is preferably equal to the number of pixel units commonly connected to one source line segment.

[0235] Of course, according to one embodiment of the present invention, the source line (S n If ) is not segmented, it is not segmented by not inserting a switch into the reference line for the same reason.

[0236] In another embodiment according to the present invention, a source line (S n Even if ) is segmented, the reference line may not be segmented. In this embodiment, between a target unit for recovering a residual value and reference units used for comparison with the residual value, a data supply line (S i , rfS iSince the number of units sharing ) changes, the parasitic capacitance changes (there are also differences in electrical characteristics due to differences in the length of the signal line used to read back the value recorded in the unit, which will be discussed later). Such differences in electrical characteristics are compensated for in the path that retrieves the pixel value recorded for reference in the reference unit. A characteristic matching circuit provided on the signal transmission path to the circuit module, which compares the pixel value retrieved from the reference unit with the residual value recovered from the target pixel unit, matches the resistance and capacitance to the target pixel unit to ensure identical electrical characteristics. This will be explained in more detail later.

[0237] In addition, in another embodiment according to the present invention, the reference unit set (200) may also be configured in the form of electrically pre-disconnected line segments without inserting a switch, as illustrated in FIG. 10. In this case, through a signal line connected to each line segment, a voltage corresponding to possible pixel values ​​(e.g., values ​​from 1 to 255) or a plurality of pixel values ​​selected among the possible pixel values ​​(e.g., a portion of possible pixel values ​​selected such that the interval between values ​​is uniform) is charged (hereinafter, the pixel value charged to the reference unit is distinguished as a 'reference value'), and also through the signal line, the current voltage of the cell capacitor of the reference unit (hereinafter referred to as the 'current reference value') is read.

[0238] The following description is based on an embodiment in which the reference unit set is configured as exemplified in FIG. 11a. Although the example in FIG. 11a shows that all possible pixel values ​​(from 1 to 255) are selected as reference values, the following description can naturally be applied to an embodiment in which only some of the pixel values ​​are selected and used as reference values.

[0239] Each reference line (rfS iEach segment of ) has a draw line (rfrL i_S[j] , i=1,.,255, j=1,..,GN, where GN is the number of segments belonging to one reference line.) is connected. Hereinafter, a set of pull lines connected to segments of the same order from different reference lines is referred to as a ‘pulling line group’.

[0240] A method for obtaining a pixel value that was applied immediately before, that is, one frame before, to an arbitrary pixel unit using the reference unit set (200) of the configuration described above will be explained in detail below.

[0241] The display control unit (110), before rendering a series of pixel data corresponding to an arbitrary scan line (hereinafter referred to as a 'current scan line') according to the order of the raster scan method, selects possible pixel values ​​for each reference line (rfS) of the reference unit set (200). i With each applied in parallel to ), (at this time, all switches (rfSW) inserted in the reference line are made to be in a short-circuit state), each select line (rfG i By driving ) simultaneously or sequentially, a set of reference values ​​(all possible pixel values, or a selected portion of them as previously described, may become the set of reference values) to the reference unit set (200) source line (S n It is charged by the number of SNum (i.e., the number of pixel units on the scan line of the panel (120)).

[0242] Then, while detecting the current residual value of each pixel unit belonging to the current scan line to which the series of pixel data to be rendered must be applied as described above, it is simultaneously checked whether there is missing data in the series of pixel data. If there is missing data, the selection line (rfG) of the reference unit set (200) in the same order as the order on the current scan line of the pixel unit to be filled with the pixel value of that data (for convenience of explanation, this order is denoted as cK (1 ≤ cK ≤ SNum). cK By driving ) the voltages charged in each reference unit, the output line group (rfrL i_S[cG] , i=1,..,255, cG is the selected line rfG cK The values ​​are detected respectively by allowing charge to be redistributed through each draw line of the reference line of the reference line segment to which the reference units connected to the gate are connected. Of course, that selection line (rfG cK Before the voltage charged by driving ) is discharged through the withdrawal line, each reference line (rfS i Open the switches at both ends of the corresponding segment of ) to electrically isolate the corresponding reference line segment.

[0243] In embodiments where the reference line is not segmented, such segment separation operation is unnecessary. Furthermore, since the reference line is not segmented, there is no need to separately circuit a draw line to read the current reference value of the reference unit for each segment, and the reference line is used as a draw line. Accordingly, the signal line referred to as the 'draw line' below may, depending on the embodiment, refer to a signal line connected to each segment of the reference line, or the reference line (rfS i It could also refer to ).

[0244] Meanwhile, in detecting the residual value of an arbitrary pixel unit through charge redistribution, as described above, an initial voltage V of a certain magnitude Det(0)In an embodiment where the voltage is applied, the current reference value of the reference unit is detected by applying an initial voltage of the same magnitude and then redistributing the voltage. In this case, the current reference value is not discharged through the extraction line, but may be additionally charged by the applied initial voltage.

[0245] The display control unit (110) obtains a value that would have been obtained if the pixel value had been detected immediately when it was filled one frame earlier (hereinafter this value is referred to as the 'previous value') by supplementing the leakage amount of time (e.g., 1 / 60, 1 / 120 second, etc.) resulting from the elapsed time of one frame according to the current frame rate for the residual value detected from the target unit to which the missing data is to be rendered. This leakage amount supplementation is performed by adding a restoration ratio k to the detected residual value, which is calculated by the following equation [5]. RES It corresponds to the operation of multiplying or amplifying.

[0246] Equation [5]

[0247] Here, T fR is the time based on the current frame rate, and T CONST is the time constant due to the closed loop of the cell capacitor and parasitic resistance of the pixel unit (or reference unit).

[0248] The above display control unit (110) is the restoration ratio k RES The above time constant T used in the calculation CONST This can be determined through testing when the display device (100) is powered on. This will be explained in more detail with an example.

[0249] When the test starts, the display control unit (110) is at the last reference line (rfS) of the reference unit set (200). 255 With a test value (e.g., 255) applied for ), the first and last selection lines (rfG1, rfG SNumAfter driving ) to charge the first and last reference units on that reference line to their respective test values, immediately that reference line (rfS 255 After opening all switches inserted in ), drive the first select line (rfG1) to the reference line (rfS 255 The first outgoing line (rfrL) connected to ) 255_S[1] The value charged in the first reference unit is primarily detected through ).

[0250] And, for a certain period of time, for example, half the time of the current frame rate, i.e., T fR After / 2 has elapsed, the last selected line (rfG SNum By driving ) the value remaining in the last reference unit the last fetch line (rfrL 255_S[GN] It is detected secondarily through ).

[0251] When the two values, rV1 and rV2, detected in the first and second stages respectively are obtained in this way, the display control unit (110) obtains the two detected values ​​and the elapsed time T fR Using / 2, the above time constant T CONST Determine according to the following equation [6].

[0252] Equation [6]

[0253] The above display control unit (110) thus has a time constant T CONST Once the determined time constant is determined, apply it to the preceding equation [5] to restore the above-mentioned k RES Determine and set the determined restoration ratio. For example, the determined restoration ratio k RES Set it as the multiplication factor of the equipped arithmetic unit or the amplification factor of the equipped amplifier.

[0254] In one embodiment according to the present invention, the operation of charging / detecting test values ​​for the reference unit set (200) as described above and obtaining the restoration ratio through this and setting it as a multiplication coefficient or amplification coefficient is performed every time between the time immediately after rendering of an arbitrary image frame on the panel (120) and immediately before rendering of the next image frame, so that electrical characteristics according to changes in the operating environment of the display panel (120) are dynamically reflected in the restoration ratio and applied to the multiplication or amplification of residual values ​​detected from the pixel units.

[0255] The above display control unit (110) is the restoration ratio k RES By applying the above-mentioned previous value obtained from the target unit, the selection line (rfG) corresponding in order in the reference unit set (200), as previously explained. cK By driving ), the corresponding draw line group (rfrL i_S[cG] Each of the series of values ​​detected through each line (i=1,..,255) is compared in parallel or sequentially.

[0256] In the description of the embodiments so far, the previous value recorded in the pixel unit is the signal line that drives the pixel value, i.e., the source line (S n The recovery line (rL) connected to ) n_S[k] It was presupposed that reading is performed through ). In accordance with the premise of such an embodiment, the current reference value extraction from the reference unit is also, in the same manner as the pixel unit, the reference line (rfS) where the reference value is driven. i The withdrawal line (rfrL) connected to ) i (In some embodiments, a reference unit set (200) is configured to be read through a reference line as illustrated in FIG. 11a.)

[0257] However, the concept and technical idea of ​​the present invention do not require that a signal line for reading a value recorded from a pixel unit and a reference unit must be connected to a supply line (source line, reference line) that drives the pixel value to the unit, or that the supply line itself. If there is another path to read the pixel value recorded in the unit, the previously recorded value may be recovered using that path.

[0258] FIG. 11b is an example of such an embodiment in which a display panel (120) is a pixel unit (OCC) as an organic light-emitting diode (OLED). n,m This concerns the case where it constitutes ).

[0259] Pixel unit (OCC) exemplified in FIG. 11b n,m In the case of ), the supply line (S) where the pixel value is driven n Through the cell capacitor (C) C In addition to the path (p10) where pixel values ​​are recorded in ), immediately before recording each pixel value, the corresponding cell capacitor (C C An initialization path (p11) is separately formed to record an initial value in ). As such, the cell capacitor (C C In cases where there are multiple paths supplying a signal for recording to ), a retrieval line (rL) for reading the previous value recorded in the corresponding unit n_S[k] ) is connected to a signal line with smaller parasitic capacitance present in the path. In the example of FIG. 11b, the supply line (S n ) has 3 transistors, and an initialization line (VL) for supplying initial values. n Since there is 1 transistor in ), the initial line (VL n A relatively smaller number of transistors exist in the path (p20) through ). Accordingly, the recovery line (rL n_S[k] ) is the initial line (VL n It is connected to ).

[0260] As exemplified in FIG. 11b, the return line (rL) is on a signal line other than the source line where the pixel value is driven. n_S[k] In an embodiment where ) is connected, it should be understood that all the descriptions regarding the source line described above apply to that other signal line. For example, instead of a source line, an initial line (VL n ) must be segmented according to this embodiment, switches are inserted between the segments, and opening / closing lines for controlling the On / Off of these switches must be connected to each, and a recovery line (rL n_S[k] ) must be connected to each segment. This is also the same in the description of the following embodiments.

[0261] And, the reference unit is also, of course, the pixel unit (OCC) of FIG. 11b. n,m It must be formed with the same structure as ), and the reference value recorded in any reference unit through the driving of the reference line is also the initial line (VL k The withdrawal line (rfrL) connected to ) i A reference unit set is configured in a form that reads through ) (depending on the embodiment, through an initial line).

[0262] FIGS. 12a and 12b respectively illustrate the relevant configuration of a comparison estimation module provided in the display control unit (110), which compares a residual value detected from a pixel unit with a current reference value detected from reference units. The reference unit set (200) described above is also included as a component of the comparison estimation module.

[0263] FIG. 12a is an example of a configuration when provided individually for each number of pixel units on the scan line of the panel (120), and FIG. 12b is an example of a configuration when provided collectively for the pixel units on the scan line of the panel (120). Accordingly, in an embodiment in which the display control unit (110) has a comparison estimation module including the configuration exemplified in FIG. 12a, a series of values ​​(val_rfrL) detected through each extraction line i_S[cG] , i=1,..,255 ) and all pixel units on the current scan line are compared simultaneously with the previous values ​​in parallel. In an embodiment equipped with a comparison estimation module including the configuration exemplified in FIG. 12b, all pixel units on the current scan line are compared sequentially with the previous values. To this end, the display control unit (110) applies a cell selection signal (px_sel) to an input switch (221) and an output switch (222) provided in the comparison estimation module according to the order of the pixel units to be compared on the scan line, thereby forming a signal path for the pixel units to be compared.

[0264] In the embodiments illustrated in FIG. 12a and 12b, the comparison / estimation unit (210), which is an internal component of the comparison estimation module n ,210) is a residual value (V) detected from the target unit in the current scan line. n,m In ) amplifier (or multiplier )(212 n In the above restoration ratio k ( ,212) RES The value amplified or multiplied by and the series of values ​​detected through each draw line ( val_rfrL i_S[cG] When compared with , i=1,..,255 ), which value among the series of current reference values ​​detected from the reference units is the previous value (V n,m Determines whether it is the closest to ).

[0265] Then, after verifying the extraction line where the current reference value identified as the closest in this way was obtained, the pixel value applied to the reference line corresponding to that extraction line is estimated as the pixel value applied to the target unit one frame prior. The estimated value (eV PRE(n,m) Prints ).

[0266] For example, the above-detected series of current reference values ​​(val_rfrL i_S[cG] , i=1,..,255 ) among val_rfrL 123_S[cG] If the value is the value closest to the previous value detected and supplemented (i.e., amplified or multiplied) by the above target unit, then the value is the detected extraction line (rfrL 123_S[cG] The reference line (rfS) connected to ) 123 The reference value 123 (or the corresponding voltage) applied to ) is estimated as the pixel value applied to the target unit one frame earlier.

[0267] FIG. 12c is a comparison / estimation unit (210) used to estimate the previous value of the corresponding pixel unit by comparing the recovered residual value with a series of reference values ​​drawn from the reference units. n, A comparison unit (2100) according to one embodiment of the present invention implemented within 210) k As an example of the configuration of ), this comparison unit (2100 k ) is, 2 comparators (2101 k ,2102 k ) and one XOR logic element (2103 k It is composed of ) and is implemented in parallel according to the number of reference values ​​to be compared (in the examples of FIGs. 11a, 12a, and 12b, since k is 255, 255 are configured in parallel within the comparison / estimation unit. )

[0268] The above two comparators (2101 k ,2102 k In ), the size of the reference value being recorded is a number of draw lines (rfrL) connected to adjacent reference units. k_S[cG], rfrLk+1_S[cG] The current reference value obtained through ) is applied to each and compared with the residual value (60) read from the target unit through the recovery line.

[0269] Therefore, multiple comparison units (2100 k In the comparison block of , k=1,..,255), in only one comparison unit, the applied residual value (60) becomes between the current two reference values ​​being compared, and the XOR logic element (2103) of this comparison unit M Only ) becomes a HIGH output state. The remaining comparison units (2103 i In , i), since the current reference values ​​applied are both smaller or both larger than the recovered residual values ​​(60), the output of the XOR logic element of the corresponding unit is LOW.

[0270] Therefore, each comparison unit (2100 i Output signal (Val) of , i=..,k,k+1,..) SEL_i The comparison / estimation unit (210) is configured in a circuit to enable the output of a device, such as a flip-flop array, that stores a corresponding reference value (i.e., a reference value recorded in a reference unit to which a value is applied through a draw line to the comparison unit), thereby enabling the output of the device. n As described above, the comparison / estimation unit (210) immediately estimates the pixel value that was immediately applied to the target unit from the simultaneous comparison of the recovered residual value (60) and a series of reference values. In another embodiment according to the present invention, without the process of selecting an extraction line from which a value closest to the previous value of the target unit is extracted and identifying a reference line corresponding to that extraction line, the comparison / estimation unit (210) n ,210) is the estimated pixel value (eV) for the target unit, which is the detection value from the extraction line closest to the previous value of the target unit. PRE(n,m) It can also be output as ).

[0271] The above comparison / estimation unit (210 nThe value output according to the comparison result in (or the value output according to the signal path selected by the output switch (222)) is applied as an estimated pixel value to the pixel data preservation unit (111) in the processing module of FIG. 7a for processing pixel data for the corresponding pixel unit, thereby causing the pixel data preservation unit (111) to use the pixel value (71) estimated according to the comparison result as described above to preserve the original information lost for the approximated missing data, that is, to construct pixel data for the missing data, or substitute data filled with some discarded LSB (or the lost value as padding of 0 or 1).

[0272] In this way, after all approximated missing data in a series of pixel data belonging to a slanted scan line is preserved, and the non-approximated pixel data is as is, each source line (S) by each of the above data driving units (113) i In a state where it is loaded on , i=1,..,SNum), by driving the gate line assigned to the current scan line, the pixel values ​​of the pixel data to be rendered on the scan line are charged into the pixel units belonging to the scan line, thereby causing the pixel data to be visually displayed on the panel (120).

[0273] As previously mentioned, when configuring a set of reference units, reference units may be assigned for all possible values ​​that a pixel value can have, and all possible pixel values ​​retrieved from the reference units may be compared with previous values ​​of the target pixel; however, in another embodiment according to the present invention, reference units may be provided for only a portion of the values ​​that a pixel can have.

[0274] For example, in the case where pixel data consists of 8 bits, reference units may be provided to provide reference values ​​for sets of pixel values ​​such as 4, 8, 16, ..., 252 or sets of pixel values ​​such as 8, 16, 24, ..., 248. When the reference values ​​are reduced in this way, the set of reference units as well as the comparison / estimation unit (210) of FIG. 12a or 12b n The circuit of (210) can be simplified. When estimating the pixel value previously applied to the target unit from the reference value, if the reference value is reduced in this way, the estimated previous value output by FIG. 12a or 12b becomes one of the reduced reference values, and thus there may be a difference from the actual pixel value previously applied to the target unit. However, the pixel data currently applied to the target unit, where the estimated previous value is used wholly or partially, is missing data, and since the approximation applied is within a range where image quality degradation as described above is not visually perceptible, using a roughly estimated previous value based on a limited number of reference values ​​does not lead to a significant degradation of image quality. However, if the number of reference values ​​is reduced too much, the possibility of a negative impact on image quality increases accordingly; therefore, the circuit configuration of the reference unit set and the comparison estimation module can be further simplified by reducing the number of reference values ​​to an appropriate number that does not affect image quality.

[0275] And, in the above-described embodiment, when providing a reference unit set, as illustrated in FIG. 11a, reference units may be provided by including as many reference units as the number of pixel units (SNum) belonging to one scan line for one reference value; however, in another embodiment according to the present invention, a number smaller than the number of pixel units (SNum) belonging to one scan line for one reference value (1 <= rUNp rVal<SNum인 경우 ) 만큼의 화소 유닛들이 구비되도록 할 수도 있다. 본 실시예에서는, 화소 유닛의 이전 값 추정을 위해서 일부의 화소 유닛에 대해서는 참조유닛 세트를 공통적으로 이용하게 된다. 즉, SNum을 rUNp rVal For groups of pixel units divided by the number, a single set of reference units is commonly used.

[0276] As a more specific example of implementation, the above rUNp rVal 1 can be applied to . This means that, in the example of FIG. 11a, only one reference unit block (201) is provided for a set of possible pixel values ​​(or a selected subset of pixel values). In this case, naturally, all pixel units belonging to the same scan line are used in common when comparing reference values ​​for estimating the previous value of a pixel unit. That is, in the case of the embodiment of FIG. 12a, the draw line (rfrL) connected to any reference unit of the reference unit block (201) i_S[1] ) is provided with comparison / estimation units (210) equal to the number of pixel units belonging to a single scan line. i It will have a circuit configuration connected to each of , i=1,..,n,..,SNum).

[0277] FIG. 13a illustrates, according to one embodiment of the present invention, that all pixel units belonging to a single scan line can commonly use a reference unit set (200') (the above numerical value rUNp rVal This is a block diagram schematically illustrating the configuration when the above comparison estimation module is implemented (when α is 1).

[0278] Each pixel column (1300) exemplified in FIG. 13a n) is illustrated on the premise that a single source line is not segmented. As previously mentioned, in the case of an embodiment in which a source line is segmented, a plurality of recovery lines connected to each of the segmented lines are connected to the input terminal of a segment group selection switch, and a single output of the segment group selection switch is each comparison / estimation unit (210 i , i=..,n-1,..,n+2,..) will be configured to be applied.

[0279] The example illustrated in FIG. 13a is for pixel units on a scan line with scan order k, residual values ​​(rmV) read through a recovery line (rLi, i=..,n,n+1,..). i,k , i=..,n,n+1,..) and each reference unit (2011) of the reference unit set (200') k For ), the withdrawal line (rfrL j A set of current reference values ​​read through , j=1,..,GN, where GN is a number less than or equal to 255), comparison / estimation units (210) provided corresponding to each column i , i=..,n-1,..,n+2,..) is compared and estimated in the manner described above, and the previous value (eV) recorded in each pixel unit at the preceding time point, corresponding to the residual value, is obtained. PRE(i,k) Shows that it determines and outputs , i=..,n-1,..,n+2,..).

[0280] The above comparison / estimation parts (210 i , i=..,n-1,..,n+2,..) can be implemented at the top or bottom of the corresponding column in correspondence with each pixel column.

[0281] As in FIG. 13a, a single reference value is compared / estimated by a large number of comparison / estimation units (210 i If configured to be used commonly by ), a relatively very large load is applied to the reference value compared to the residual value. The difference in relative load is, each comparison / estimation unit (210 iSince the difference in the arrival time of the residual value and the reference value compared to ) and the difference in voltage attenuation can lead to errors in comparison, a driver such as a source-follower that drives common connected signal lines with the reference value read from the reference unit is provided to reduce the load on the reference value, as exemplified in FIG. 13a.

[0282] In addition, each recovery line (rL i In ), the reference value is the corresponding comparison / estimation part (210 i A circuit element for signal delay may be inserted so as to be synchronized with the time of reaching ). The circuit element inserted for this purpose also corresponds to the compensation circuit inserted into the recovery line to match the electrical characteristics between the recovery line and the withdrawal line, as described later.

[0283] Even if the reference value read from the reference unit set (200') is driven using an actuator, each comparison / estimation unit (210 i If the path length to the input terminal of ) increases, signal loss due to path resistance may increase, which may lead to inaccurate comparison of magnitudes. Considering this point, each comparison / estimation unit (210 i It is desirable to implement a circuit for providing reference values ​​such that the path length of the reference value input to ) is within an appropriate range.

[0284] FIG. 13b conceptually illustrates the layout on the display panel (120) of a comparison estimation module according to one embodiment of the present invention, taking these points into consideration.

[0285] In the embodiment illustrated in FIG. 13b, a set of reference units (200') is positioned in the center of a display panel (120), and a drawing line (rfrL) from each reference unit of the set (200') L) branches off from the central part and is configured so that the signal is driven to both sides by two drivers (231). And, both drivers (231) are provided with comparison / estimation units (210) corresponding to each pixel column. i It is configured to divide the reference value applied to itself into two and supply it in common. When configured to supply the reference value in this way, the average length of the reference value supply path is shortened to half compared to when it is not.

[0286] In the same way, if the reference unit sets are further distributed evenly on the display panel (120) and the extraction lines from each reference unit set are branched to be driven in both directions by both drivers, the average length of the reference value supply path can be reduced more significantly. For example, four reference unit sets (the aforementioned numerical value rUNp rVal When the value is 4, if the values ​​are distributed evenly across the panel area, the average length of the reference value supply path is shortened to 1 / 8.

[0287] Meanwhile, as described above, the method of estimating the pixel value that was recorded immediately before for a pixel unit by recording a reference value in reference units and then reading the current reference value from the reference unit and comparing it with the residual value of the pixel unit is such that even if the operating environment of the panel (120) changes, the electrical influence according to the changed environment acts equally on the pixel unit and the reference unit, so the relative magnitude of the values ​​detected from both units does not change even with changes in the operating environment. Therefore, the pixel value applied to the previous frame for any pixel unit can always be obtained accurately regardless of changes in the operating environment.

[0288] Each recovery line (rL) provided to detect the residual value from the pixel units on the panel (120) above. n_S[i], i=I,..,k,..) has different lengths to the outer edge of the panel (120) where the display control unit (110) is physically located, for example, to the top or bottom. This means that even if the same pixel value is recorded for the top pixel unit and the bottom pixel unit, the electrical characteristics, such as parasitic resistance of the corresponding recovery line, are different, so the residual value may be detected as a different size when detected through the corresponding recovery line.

[0289] In addition, the extraction line for reading the current reference value from each reference unit of the reference unit set is inevitably different from the retrieval line for reading the residual value of the pixel unit in terms of distance to the comparison estimation module exemplified in FIG. 12a or 12b for comparing the detected residual value with the current reference value. Due to this difference or other factors (parasitic characteristics of commonly connected circuit elements, line layout, etc.), there may also be differences in the electrical characteristics (impedance, signal delay characteristics, etc.) acting on the line. These differences in electrical characteristics cause a difference when detected by charge redistribution, even if the residual value of the pixel unit and the reference unit and the current reference value are actually identical. Furthermore, they cause a time difference in the time it takes for the residual value detected by charge redistribution and the current reference value to reach the comparison estimation module, respectively. This phenomenon leads to inaccuracy in determining the previous value of an arbitrary pixel unit through comparison with the reference value.

[0290] Accordingly, in one embodiment according to the present invention, in order to prevent electrical characteristics caused by differences such as lengths between these recovery lines from affecting the comparison of the detected residual value with the current reference value, and also to prevent differences in electrical characteristics between the recovery line and the withdrawal line from affecting the comparison of both values, the display control unit (110) includes a line characteristic changing unit (300) for changing the electrical characteristics of the withdrawal line for detecting the current reference value from the reference unit set, configured as illustrated in FIG. 14, and the withdrawal lines (rfrL) for detecting the voltage of the reference unit i_S[k] , i corresponds individually to all or part of {1, 2, .., 255} and is provided for each of ).

[0291] The above line characteristic changing unit (300) comprises circuit elements (320) that exhibit different electrical characteristics (impedance, capacitance, or signal delay characteristics, etc.), as illustrated. i , i=..,k,..)( semiconductor devices such as resistors, capacitors, or MOSFETs, etc.) inserted into circuit lines (eqL i A line characteristic compensation unit (320) in which , i=1,2,..,GN) are arranged in parallel, and the circuit lines (eqL i A first channel switch (310) that selects one of , i=1,2,..,GN) and connects it to an output terminal, and one extraction line, the circuit lines (eqL i It is configured to include a second channel switch (330) that connects to a selected circuit line among , i=1,2,..,GN).

[0292] In the line characteristic compensation unit (320) configured according to one embodiment of the present invention, the number of circuit lines provided GN is equal to the number of line segments formed by dividing any one source line with a switch (or in a form that is disconnected from each other beforehand) and, of course, the number of recovery lines connected to each segment of any one source line.

[0293] And, each circuit line (eqL i Circuit elements (320) inserted in , i=1,..,k,..GN) i , i=..,k,..) is such that the electrical characteristics of the entire signal line formed by the corresponding circuit line together with the extraction line and switches (310, 330) are any one source line (S n In ), the corresponding circuit line in order (eqL i A return line (rL) connected to a line segment in the same order as ) n_S[k] It is equipped to have electrical characteristics that compensate so as not to differ from the electrical characteristics of the entire path by ).

[0294] The configuration of the line characteristic change unit (300) for matching the electrical characteristics of the withdrawal line to the recovery line, including line characteristic compensation units of different electrical characteristics equal to the number of recovery lines, is merely a simple example of implementing the principles and technical ideas of the present invention, and the electrical characteristics can be matched with each recovery line using line characteristic change units that are not limited to the same number. For example, the line characteristic change unit may be configured by replacing the line characteristic compensation unit (320) with a resistance / capacitance load modeling circuit. In this case, when the line characteristic change unit is configured with a resistance / capacitance load modeling circuit, each characteristic matching output point of the modeling circuit is connected to the output terminal switch (310).

[0295] Meanwhile, when the comparison estimation module exemplified in FIG. 12a or 12b is placed at any position on the panel (110), the length from the recovery line connected to the corresponding line segment to the comparison estimation module for detecting residual values ​​from each pixel unit on any scan line also differs, and this difference in length may also appear as a difference in electrical characteristics. Therefore, taking this into consideration, the line characteristic compensation unit (320) configured as exemplified in FIG. 14, which is inserted into the corresponding extraction line according to the extraction line connected to the reference unit for comparison with the residual value of a pixel unit in a certain order on any scan line, may have different electrical characteristics from the line characteristic compensation unit of the same configuration inserted into other extraction lines and the circuit elements inserted to compensate for electrical differences in each circuit line. The same applies to an embodiment in which the line characteristic compensation unit is configured as a resistance / capacitance load modeling circuit.

[0296] As a more specific example, the draw line rfrL forms a connection path from a reference unit to a comparison estimation module for comparison with the residual value of the first-order pixel unit of any scan line belonging to the ln-th line segment. i_S[1] In a line characteristic compensation unit inserted and provided in ( i=1,2,..,255, or a part selected therefrom), circuit line eqL corresponding to the ln-th recovery line ln Circuit elements of C_Comps ln is an rfrL extraction line that forms a connection path from a reference unit to a comparison estimation module for comparison with the residual value of the pixel unit in the last sequence of the same scan line. i_S[GN] In the line characteristic compensation unit inserted therein, the circuit line eqL corresponding to the ln-th recovery line ln Circuit elements of C_Comps lnBy having different electrical characteristics from each other, the difference in electrical characteristics with the corresponding return line connected to the pixel units on the same scan line is compensated accordingly.

[0297] According to another embodiment of the present invention, the reference unit set of FIG. 11a may be provided for each pixel row (i.e., scan line) of the panel (120). In such an embodiment where the reference unit set (200) is provided for each pixel row, the length of the recovery line to the comparison estimation module provided in each column of FIG. 12a becomes equal to the extraction line regardless of the position of the pixel unit on the scan line. Therefore, in this embodiment, unlike the previous embodiment, the line characteristic compensation unit (320) of the line characteristic change unit (300) of FIG. 14 does not need to have electrical characteristics to compensate for the difference in the length of the recovery line according to the position of the pixel unit on the scan line.

[0298] As explained above, the current reference value (val_rfrL) that is finally detected through the line characteristic changing unit (300) i_S[k] ) is any one source line (S n The recovery line (rL) connected to the line segment of the corresponding order (k-th) in ) n_S[k] When detecting through ), the residual potential value of the pixel unit is detected while being affected by the same electrical influence as the characteristics of the extraction line. Therefore, if the residual value of any pixel unit is the same as any current reference value, then even if the detection of that value is affected by any electrical influence, the two values ​​being compared after passing through the line characteristic changing unit (300) will be determined to be the same.

[0299] In an embodiment in which the line characteristic compensation unit (320) configured as illustrated in FIG. 14 is provided for each extraction line, the display control unit (110) connects to the next sequence of circuit lines through a segment selection signal (sel_sect) applied to the first channel switch (310) and the second channel switch (330) respectively for each line characteristic compensation unit whenever the line segment to which the pixel units to which the residual value is to be detected changes.

[0300] A reference value reflecting the same characteristics as the influence of electrical characteristics received by the residual value detected through the recovery line, which is output by each line characteristic compensation unit, is a comparison / estimation unit (210) of FIG. 12a or 12b n It is applied to ,210) and, as described above, is compared with the residual value detected from the pixel unit through the recovery line, and is used to estimate the pixel value that was applied to the pixel unit in the previous frame.

[0301] In the above-described embodiment, the difference in electrical characteristics between the recovery line connected to detect the residual value of the pixel unit and the extraction line connected to read the current reference value of the reference unit was offset through a line characteristic changing unit (300) configured as exemplified in FIG. 14, which is inserted into the extraction line. However, there may be cases where the difference in electrical characteristics between the two signal lines of the recovery line and the extraction line cannot be completely eliminated by this embodiment alone, and there may be a residual characteristic difference that is not offset by the insertion of circuit elements into the extraction line to offset the electrical characteristics.

[0302] For example, there may be cases where the return line from the pixel unit is shorter than the draw line from the reference line, and there may also be cases where the difference in capacitance additionally generated by inserting a semiconductor device such as a MOSFET into the draw line to offset the signal delay characteristics on the return line must be added to the return line.

[0303] Accordingly, regarding a specific embodiment of the present invention, it has been described that a compensation circuit element is inserted into the extraction line to offset the difference in electrical characteristics between the recovery line and the extraction line, which are respectively connected to the unit (more specifically, to the line segment to which the unit is connected), for the purpose of comparing the residual value of the pixel unit with the current reference value of the reference unit; however, a method of offsetting the difference in electrical characteristics between the two signal lines through a compensation circuit inserted into the recovery line connected to the pixel unit can also be naturally applied as an embodiment of the present invention. Furthermore, a method of offsetting the difference in electrical characteristics between the two signal lines by inserting compensation circuits with different electrical characteristics into each of the two signal lines can also be applied as an embodiment of the present invention.

[0304] Up to now, embodiments according to the present invention have been described on the premise that the pixel unit (and reference unit) includes a cell capacitor. That is, all embodiments according to the present invention are based on the premise that a voltage corresponding to the color (or luminance) to be expressed is charged to the cell capacitor.

[0305] However, the present invention does not presuppose how the voltage charged in the cell capacitor is expressed as color. Therefore, the various embodiments of the present invention described above can naturally be applied as is to LCD panels that express color in pixel units by controlling the amount of light passing through a color filter by the voltage charged in the cell capacitor, OLED panels that express color in pixel units by controlling the amount of light emitted by a direct light-emitting element by the voltage charged in the cell capacitor.

[0306] Ultimately, regardless of the name used for the method of expressing color, if the panel is a display type in which the voltage required for the expression of color (or luminance) is charged to the cell capacitor included in the pixel unit and the charged voltage is used for color expression, the concept and technical principles of the present invention can be applied as is.

[0307] Therefore, if a color (luminance) expression unit element is configured to include a capacitor that charges a voltage for the expression of color or luminance, regardless of the name used for it, it should be interpreted as being encompassed within the scope of the meaning of the term pixel (or reference) unit used in the specification and claims of the present invention.

[0308] In addition, the pixel data approximation method of the various embodiments described so far may also be applied to data that corrects the pixel data originally contained within the input or received image. For example, the pixel data of each image block within the image may first be corrected through modeling that adapts the image displayed by the display device to the viewing environment and the characteristics of the corresponding image block, and then the approximation may be applied to each modeled image block.

[0309] In the modeling applied at this time, corrections are used to change the pixel data of the corresponding image block to a lower brightness value when the brightness around the device is darker, to a lower brightness value when the screen reflectivity of the display device is high, or to change the pixel data of the corresponding target block to a higher brightness value when the brightness set on the display device is high. The modeling using these corrections corresponds to a color space conversion process that changes the pixel data into what the viewer actually perceives visually when the corresponding image block is displayed on the display device.

[0310] Various embodiments of the apparatus and method for approximating pixel data and reusing it on a screen according to the present invention, as well as the structures and operations described in those embodiments, can be selectively combined in various ways, provided that they are not mutually incompatible.

[0311] The embodiments of the present invention described above are disclosed for illustrative purposes only, and those skilled in the art may modify, change, substitute, or add various other embodiments within the technical concept and scope of the present invention as disclosed in the appended claims.

Claims

1. A device for rendering an input image on a screen, It is configured to determine the characteristics of pixel data within an image block of a predetermined size, or the characteristics of an area designated for said pixel data, and to determine whether to approximate said pixel data based on said characteristics, and additionally, for pixel data for which approximation is determined within said image block, from the lower N ( 0 <N<=픽셀 데이터의 비트폭 )비트의 값을 버린 결손 데이터로 만드는 근사화를 적용하도록 구성된 근사화부; A buffer configured to store together the missing data to which the approximation by the approximation unit is applied and the pixel data to which the approximation is not applied; and A device comprising a control unit that receives data stored in the buffer from a data supply unit that reads data stored in the buffer according to a specified timing, and causes the data to be visually displayed on a provided display panel.

2. In Paragraph 1, The pixel data to which the application of the above approximation is determined is one in which the brightness is less than or equal to a predetermined lower limit or greater than or equal to a predetermined upper limit, and The approximation unit determines N to a larger value in proportion to the degree to which the brightness of the pixel data to which the approximation is applied is separated below the lower limit or above the upper limit, A device for determining the N for the second pixel data to be a larger value than the N for the first pixel data when the degree to which the brightness of the first pixel data in the image block is separated from the upper limit and the degree to which the brightness of the second pixel data in the image block is separated from the upper limit are the same.

3. In Paragraph 1, The approximation unit determines whether to apply the approximation to the pixel data based on the distribution of the hardness characteristic values ​​of the pixel data within the designated area, and The above-mentioned hardness characteristic value is the ratio of the value reflecting the difference between pixel values ​​of any two pixel data to the value reflecting the spatial distance between those two pixel data, and The approximation unit compares the distribution of the hardness characteristic values ​​of the pixel data of the specified area with a plurality of reference distributions to identify one reference distribution that is closest to the distribution of the hardness characteristic values, and determines a value specified for the identified reference distribution as the N for the pixel data for which the approximation is determined.

4. In Paragraph 3, The above approximation unit is a device configured to determine N as a larger value when the distribution of hardness characteristic values ​​of pixel data within the specified area is skewed toward a high hardness characteristic value compared to when it is skewed toward a low hardness characteristic value.

5. In Paragraph 1, The approximation unit applies the approximation to the arbitrary pixel data when the arbitrary pixel data differs from the average pixel value of the pixel data within the designated area by more than a predetermined threshold value. A device configured to determine N as a larger value when the difference from the reference value of the pixel data based on the average pixel value is large compared to when it is small.

6. A device for rendering an input image on a screen, It is configured to determine the characteristics of pixel data within an image block of a predetermined size, or the characteristics of an area designated for said pixel data, and to determine whether to approximate said pixel data based on said characteristics, and additionally, for pixel data for which approximation is determined within said image block, from the lower N ( 0 <N<=픽셀 데이터의 비트폭 )비트의 값을 버린 결손 데이터로 만드는 근사화를 적용하도록 구성된 근사화부; A buffer configured to store together the missing data to which the approximation by the approximation unit is applied and the pixel data to which the approximation is not applied; and It includes a control unit that receives data stored in the buffer from a data supply unit that reads data stored in the buffer according to a specified timing and makes it visually displayed on a provided display panel. The approximation unit is configured to determine at least one of the first, second, and third approximation methods based on the degree of brightness of the arbitrary pixel data and the corresponding degree to the edge, wherein The above first approximation method is a method for determining the number of bits to discard from the lower part according to the brightness characteristics of the given pixel data, and The above second approximation method is a method for determining the number of bits to discard from the lower part for the given pixel data based on the distribution of the hardness characteristic values ​​of pixel data within a region centered on the given pixel data, and The above third approximation method is a method for determining the number of bits to discard from the lower part of the given pixel data based on the degree to which the difference between the average pixel value of the area centered on the given pixel data and the given pixel data deviates from a predetermined threshold. A device in which the above-mentioned hardness characteristic value is the ratio of a value reflecting the difference between pixel values ​​of any two pixel data to a value reflecting the spatial distance between the two pixel data.

7. In Paragraph 1, The above N is a device determined by a calculation with a weight determined according to the difference between the reference contrast, which is determined in correspondence with the spatial frequency of the image block, and the contrast of the image block.

8. In Paragraph 1, The above approximation unit is further configured to change the missing data created by the approximation into missing data obtained by performing an AND operation with a representative value determined for the specific pixel value range when the pixel data belongs to a specific pixel value range, The above representative value is a device in which the pixel value with the largest number of lower consecutive zero bits among the pixel values ​​corresponding to the above specific pixel value range.

9. In Paragraph 8, The device is further configured such that the approximation unit is configured to dynamically change the setting based on the brightness around the device for at least one boundary of the specific pixel value range.

10. In Paragraph 1, The above approximation unit applies an approximation to all pixel data within the image block to produce missing data by discarding the N bit value from the lower end, The above approximation unit is configured to determine N according to the low contrast degree of the image block's contrast compared to a reference contrast determined in correspondence with the spatial frequency of the image block.

11. In Paragraph 10, The above approximation unit is configured to determine N as a value smaller than the value determined when the contrast of the image block is lower than the reference contrast when the contrast of the image block is higher than the reference contrast.

12. In Paragraph 10, The approximation unit is configured to set the reference contrast differently for each spatial frequency of the image based on at least one of the brightness around the device and the distance from the device to the viewer. If the brightness becomes brighter or the distance becomes farther, the reference contrast is set in such a way that it becomes a relatively higher contrast. The approximation unit is configured to correct the first bit number determined according to the degree of low contrast based on the difference between the first bit number and the second bit number obtained by applying an approximation method not based on the reference contrast to the pixel data, and to determine the corrected first bit number as N.

13. In Paragraph 1, The device is configured such that the approximation unit applies the approximation to the pixel data of the subsequent block when the block difference component obtained from the difference in pixel values ​​between the image block and at least some corresponding pixel data of the image block and the block that is temporally consecutively followed by the image block is less than or equal to a predetermined reference value.

14. In Paragraph 6, The above approximation unit is, Apply at least one of the fourth and fifth approximation methods, but, The above-described fourth approximation method is an approximation in which, for pixel data belonging to the image block, if the difference in pixel values ​​between the pixel data and the position-corresponding pixel data of a block that is temporally consecutively following the image block is all below a reference value, the subsequent block is not stored in the buffer. The above fifth approximation method is an approximation in which, for pixel data at the boundary of the image block, if the difference in pixel value between the pixel data and the pixel data corresponding to the position of a block that is temporally consecutively following the image block is less than or equal to a reference value for all pixel data at the boundary, the subsequent block is not stored in the buffer. A device in which the above boundary band is a boundary of the image block or a boundary band having a bit width of 2 or more.

15. In Paragraph 14, The device is configured such that, even if the pixel data of all or part of the image block is compared with the subsequent block and it is determined that the subsequent block should not be provided to the control unit, the pixel data of the subsequent block is stored in the buffer and provided to the control unit at every point in time when a predetermined period has elapsed.

16. In Paragraph 1, The control unit above, in the transmitted data, if all data to be applied to any pixel unit among a series of pixel units belonging to a scan line on the display panel to be visually manifested at the current time is missing, or from the lower K ( 0 <K<픽셀 데이터의 비트폭 )비트의 값이 버려진 부분 결손 데이터인 경우에는, 상기 임의의 화소 유닛에 현재 남아 있는 잔류값으로부터 상기 임의의 화소 유닛에 직전에 인가하였던 화소값을 구하고 그 구한 화소값 전부를, 또는 그 구한 화소값에서 상기 K 비트에 대응되는 값을 부분 결손 데이터에 결합한 화소값을 상기 임의의 화소 유닛에 인가하도록 구성된 것인 장치.

17. In Paragraph 1, A device in which the approximation for the image block or pixel data within the image block is performed by the execution of a neural network that has undergone deep learning on the approximation results of a plurality of images.

18. In Paragraph 1, The above data supply unit is configured to transmit, with respect to the approximation information for the pixel data to which the approximation is applied, to the control unit during the transmission of a series of pixel data corresponding to any scan line on the display panel to which the pixel data to which the approximation is applied belongs, or a video frame containing the series of pixel data, or such transmission, wherein The above approximation information is information indicating the number of bits discarded from the lower part of the pixel data to which the approximation is applied, or information indicating that the entire pixel data is missing due to the application of the approximation. The above data supply unit transmits approximation information for a series of pixel data corresponding to the arbitrary scan line to the control unit based on a horizontal or vertical synchronization signal of the display panel, wherein the information regarding which image block the transmitted approximation information should be applied to is not provided to the control unit.

19. In Paragraph 1, The above approximation unit is configured to store only the remaining bits excluding the N bits for the above missing data in the buffer, and also to store approximation information of the number representing the N bits in the buffer. A device in which the data supply unit is configured to fill the N bits from the lower with 0 or 1 by referring to the approximation information for the missing data and transmit it to the control unit, or the control unit is configured not to drive the pixel unit at the position where the N bits of the missing data are to be expressed.

20. In Paragraph 1, A device in which the image block of the above predetermined size is a block in which pixel data is modified by modeling the corresponding image block within the input image into a visual image perceived by the viewer in their current viewing environment.

21. In Paragraph 1, It further includes a display panel configured to include an array of pixel units that cause authorized pixel values ​​to be visually expressed, and The above control unit is configured such that when pixel data belonging to an image frame is input, it applies a pixel value corresponding to each pixel data to a pixel unit at a corresponding display position on the display panel to be visually displayed, An apparatus configured to detect a residual value currently remaining in a first pixel unit corresponding to a position where the original information is lost in the image frame, and to construct substitute data that preserves the lost information of the missing data based on the detected residual value, and then apply a pixel value corresponding to the substitute data to the first pixel unit.

22. In Paragraph 21, The above control unit is configured to form an electrical path with the first pixel unit and, depending on the formation of the electrical path, to detect the residual value based on a voltage in which the amount of charge currently remaining in the capacitor within the first pixel unit is balanced with another capacitor on the formed electrical path through charge sharing.

23. In Paragraph 21, The above control unit is configured to supplement the leakage amount during the time difference in the detected residual value based on the time difference between the time when a pixel value is immediately applied to the first pixel unit and the time when the missing data is preserved based on the residual value, and to preserve the lost information of the missing data based on the residual value in which the leakage amount has been supplemented.

24. In Paragraph 21, The display panel includes a configuration in which a signal line commonly connected to apply a specific value to a series of pixel units including the first pixel unit comprises a plurality of line segments distinguished by a plurality of switches inserted and connected within the signal line. The above control unit is configured to detect the residual value of any pixel unit after opening a switch connected to the line segment to which the first pixel unit belongs among the plurality of switches.

25. In Paragraph 24, A device in which, when there are multiple driving lines capable of applying a specific value to the first pixel unit, the signal line corresponds to the supply line with the smallest parasitic capacitance among the multiple line segments.

26. In Paragraph 21, The above-described display panel is a device having a configuration in which a series of pixel units arranged in any column are distributed among a plurality of line segments arranged in a row while being electrically disconnected from each other, and a plurality of pixel units are commonly connected to each line segment.

27. In Paragraph 24, The above line segments have a structure in which they are individually connected to the lowest signal line links of signal line links having a tree structure, starting from the outer edge of the display panel with one or more lines at the top, and a path selection switch is inserted at each point where the signal line links branch. A device in which the signal line formed by the signal line links electrically connected to each other by the path selection of each of the above path selection switches is commonly used when applying a pixel value to pixel units commonly connected to the line segment to which the signal line is connected, and when detecting a residual value currently remaining in the pixel units.

28. In Paragraph 21, The above display panel further comprises a set of reference units in which a plurality of reference units having the same circuit structure as the pixel unit are arranged, and The above control unit is configured to apply different pixel values ​​to reference units within the reference unit set, and then compare the current value detected by each reference unit of the reference unit set with the residual value, and use the pixel value applied to the reference unit where the current value closest to the residual value was detected to preserve the lost information of the missing data.

29. In Paragraph 28, The above display panel is individually provided with a plurality of circuit lines having different electrical characteristics for transmitting the current value to any reference unit within the set of reference units for comparison with the residual value. The control unit is configured to select one of the plurality of circuit lines according to which line segment the arbitrary pixel unit is connected to among a series of pixel units forming a column including the arbitrary pixel unit, and to detect the current value of each reference unit through the selected circuit line and compare it with the residual value. A device in which the above series of pixel units are distributed and arranged across a plurality of line segments, and the pixel units distributed and arranged across each line segment are commonly connected to the corresponding line segment.

30. In Paragraph 29, A device having, in at least one of a first signal line for detecting a current value from any reference unit within the set of reference units and a second signal line for detecting the residual value of any pixel unit, a circuit that compensates for the difference in electrical characteristics between the first signal line and the second signal line up to a circuit where the current value and the residual value are compared with each other.

31. In Paragraph 28, The above reference unit set is configured such that the reference units are arranged in N1 x N2 (where N1 is any number greater than or equal to 2 and less than or equal to the number of possible pixel values, and N2 is any number greater than or equal to 1 and less than or equal to the number of pixel units on the scan line of the display panel), The control unit is configured such that, for the N1 x N2 reference units, whenever a pixel value for each scan line of the display panel is applied, the reference unit set is updated in such a manner that the N1 different pixel values ​​are identically recorded in the N2 reference units for each pixel value. The above control unit is provided at the top or bottom corresponding to each pixel column of the display panel, and includes a plurality of comparison / estimation units that compare the N1 current detection values ​​with the residual value detected from the pixel unit of the corresponding column and determine the nearest current value from the comparison.

32. A device for rendering an input image on a screen, A display panel comprising an array of pixel units that cause an authorized pixel value to be visually expressed, and a set of reference units in which a plurality of reference units having the same circuit structure as the pixel units are arranged, and The device is configured to include a control unit configured such that when pixel data belonging to a video frame is input, a pixel value corresponding to each pixel data is applied to a pixel unit at a corresponding display position on the display panel to be visually displayed, wherein The control unit is configured to detect a residual value currently remaining in any pixel unit on the display panel based on a voltage generated by charge sharing in a capacitor on an electrical path formed for the any pixel unit, and to estimate a pixel value immediately recorded in the any pixel unit through comparison with reference values ​​detected in a series of reference units within the reference unit set for the detected residual value. Additionally, the device is configured such that at least one of the two signal lines to which any one of the detected reference values ​​and the detected residual value are transmitted for comparison has a compensation circuit to make the electrical characteristics of the two signal lines the same.

33. In Paragraph 31, The above display panel has a configuration in which a supply line commonly connected to apply a specific value to a series of pixel units including the arbitrary pixel unit is composed of a plurality of line segments distinguished by a plurality of switches inserted and connected within the supply line, and The above control unit is a device configured to detect the residual value of the arbitrary pixel unit after opening the switch connected to the line segment to which the arbitrary pixel unit belongs among the plurality of switches.

34. In a method for rendering an input image on a screen, Identifying the characteristics of pixel data within an image block of a predetermined size, or the characteristics of an area specified for said pixel data, and based on the identified characteristics, for said pixel data, from the bottom N ( 0 <N<=픽셀 데이터의 비트폭 )비트의 값을 버린 결손 데이터로 만드는 근사화를 적용하는 단계; A step of storing one or more missing data to which the approximation is applied and pixel data to which the approximation is not applied together in a buffer within the image; A step of reading data stored in the above buffer according to a specified timing and displaying it visually on a display panel; A step of receiving pixel data belonging to an image frame; and The method includes the step of applying a pixel value corresponding to each of the received pixel data to a pixel unit at a corresponding display position on a display panel to visually display the pixel data; A method comprising the step of visually manifesting the pixel data, wherein if the missing data in which the original information is lost is within the image frame, detecting a residual value currently remaining in an arbitrary pixel unit corresponding to a position where the missing data is to be displayed on the display panel, constructing substitute data that preserves the lost information of the missing data based on the detected residual value, and then applying a pixel value corresponding to the substitute data to the arbitrary pixel unit.

Images