LED backlight dimming method and system

Abstract

The invention provides an LED backlight dimming method and system, and belongs to the field of image data processing.The method comprises the steps that an image frame sequence used for LED screen display is obtained, a brightness highlighting channel is extracted from the image frame sequence, and backlight areas are divided for all image frames according to the brightness highlighting channel; and calculating a back moving mask for the backlight area to obtain a backlight moving area, and adjusting the backlight moving area. The method is suitable for real-time operation on embedded equipment, and the image quality and the user experience are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the field of image data processing, and in particular relates to a method and system for dimming an LED backlight source. Background Art

[0002] With the rapid development of display technology, liquid crystal displays (LCDs) have been widely used in electronic devices such as televisions, mobile phones, tablets and laptops. As an important component of LCD, the dimming technology of the backlight directly affects the display effect and energy consumption level. The traditional backlight dimming method is mainly based on global backlight control, that is, the backlight brightness of the entire screen is uniformly adjusted. However, global dimming cannot finely control the local brightness and darkness differences of the image content, resulting in the black of the display not being deep enough, the white not being bright enough, and the overall contrast being limited. Even when displaying dark scenes, the global backlight still maintains a high brightness, resulting in unnecessary energy consumption and reducing the energy efficiency of the device. In high dynamic range images, the details of the bright and dark parts cannot be fully displayed, affecting the user's visual experience.

[0003] Regional backlight dimming technology can divide the screen into multiple backlight areas, and independently adjust the backlight brightness according to the image content of each area, thereby improving contrast and reducing energy consumption. However, since the brightness change of the backlight area does not completely match the edge of the actual image content, the light in the bright area leaks into the dark area, resulting in a halo, which affects the clarity and detail of the image. Traditional dimming methods usually only consider brightness information and ignore hue and color saturation. For example, a dynamic dimming device for a liquid crystal display LED backlight source described in the patent document with publication number CN103778898B, although the dynamic dimming device can effectively solve the problem of screen flickering, it may still cause color distortion, especially when displaying low brightness or night scenes, the color accuracy is insufficient. In fast-moving pictures, the backlight brightness cannot respond to changes in image content in time, resulting in motion blur and smearing, affecting the clarity of dynamic pictures. Some advanced image segmentation and motion detection algorithms, such as convolutional neural networks based on deep learning, although they improve dimming accuracy, have large computational complexity and high hardware performance requirements, and are not suitable for resource-constrained embedded devices. There are limitations on the response speed and control accuracy of the backlight driver hardware, and complex dimming algorithms may not be able to fully exploit their advantages. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for dimming an LED backlight source to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.

[0005] The present invention proposes an improved LED backlight dimming method. The method extracts the brightness highlight channel of the image, combines the information of brightness, hue and color saturation, and uses an image segmentation method based on wavelet analysis to divide the image into several backlight areas. Subsequently, the backlight motion area is quickly and accurately identified. The brightness of the motion backlight area is finely adjusted, and the halo effect is minimized to reduce brightness leakage. The whole process has low computational complexity and is suitable for real-time operation on embedded devices, which significantly improves image quality and user experience.

[0006] In order to achieve the above object, according to one aspect of the present invention, a method for dimming an LED backlight source is provided, the method comprising the following steps: Acquire an image frame sequence for LED screen display, extract a brightness highlight channel from the image frame sequence, and divide each image frame into backlight areas according to the brightness highlight channel; Calculate the backlight motion mask for the backlight area to obtain the backlight motion area; Adjust the backlight motion zone.

[0007] Furthermore, the image output after the backlight motion area is adjusted is processed to minimize the halo effect.

[0008] Furthermore, the method for dividing each image frame into backlight areas according to the brightness highlight channel is: using an image segmentation method based on wavelet analysis and wavelet transform to perform image segmentation and obtain a number of backlight areas.

[0009] Furthermore, the image frame sequence used for LED screen display is a time series of an image matrix, which is composed of pixel values ​​of each row and column. Brightness, hue and color saturation are separated from the pixel values, and a brightness highlighting channel is obtained by combining the contrast of brightness with hue and color saturation.

[0010] Furthermore, the backlight motion mask is calculated for the backlight area, and the method for obtaining the backlight motion area is specifically as follows: For the backlight area, the absolute value of the difference between the motion frame path and the motion backlight path is calculated as the backlight motion mask of the backlight area, and whether the backlight area is a backlight motion area is determined by the backlight motion mask.

[0011] Preferably, it is determined whether the backlight motion mask is not less than 1: if yes, the backlight area is considered as a backlight motion area; if no, the backlight area is considered as a static area. Here, the absolute value of the cosine similarity in the positive and negative intervals is 1. Once it is greater than 1, it indicates that a crossover in the linear probability direction has occurred, so that the area with significant changes can be identified as a motion area. If the brightness change of the backlight area is significantly different from the overall change, it means that the area has independent motion and needs special processing.

[0012] The motion frame path can be used to represent the overall degree of motion between frames based on the brightness highlight channel matrix of the current frame and the previous frame, and the motion backlight path can be used to represent the degree of motion of the backlight area relative to the global state based on the brightness highlight channel data in the backlight area and the brightness data at the corresponding position on the global diagonal line.

[0013] Among them, for calculating the motion frame path, an optimal method is provided: extract the brightness highlight channel matrix of the current frame and the previous frame, perform feature extraction on the brightness highlight channel matrix of the current frame and the previous frame, and obtain their respective feature vectors; calculate the cosine similarity between the feature vector of the current frame and the feature vector of the previous frame as the motion frame path.

[0014] Among them, for calculating the moving backlight path, a preferred method is provided: for each backlight area in the current frame, the geometric center position of the backlight area is calculated; in the current frame, the brightness highlight channel data on the diagonal is obtained to form a global diagonal array; through the geometric center point of the backlight area, a straight line parallel to the global diagonal is drawn, and the brightness highlight channel data located on the parallel straight line in the backlight area is extracted to form a regional brightness array; the parallel straight line is translated or projected onto the global diagonal, and the brightness highlight channel data at the corresponding position is obtained to form an aligned brightness array; the cosine similarity between the regional brightness array and the aligned brightness array is calculated as the moving backlight path of the backlight area.

[0015] Furthermore, the method for adjusting the backlight motion area is: The brightness value of each pixel in the backlight motion area is adjusted by dividing the brightness value of the backlight motion mask of the backlight area.

[0016] Furthermore, the method for adjusting the backlight motion area may be: If a backlight area in the image frame currently displayed on the screen is determined to be a backlight motion area, the values ​​of the brightness highlight channel corresponding to each pixel point in the backlight motion area are respectively divided by the motion frame path of the image frame currently displayed on the screen relative to the previous image frame, and the obtained quotients are respectively used as the backlight motion sub-masks corresponding to each pixel point in the backlight motion area; The brightness value of each pixel in the backlight motion area is divided by the backlight motion sub-mask corresponding to each pixel to obtain the brightness adjusted by the backlight motion sub-mask for output on the LED screen, thereby adjusting the brightness value on the image frame output on the LED screen.

[0017] The present invention also provides a system for dimming an LED backlight source, the system comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, the processor implementing the steps in the method for dimming an LED backlight source when executing the computer program, the system can be run on computing devices such as desktop computers, notebook computers, mobile phones, PDAs, and cloud data centers, and the executable system may include, but is not limited to, a processor, a memory, and a server cluster, and the processor executes the computer program to run in the following system units: A data extraction unit, used to obtain an image frame sequence for LED screen display, extract a brightness highlight channel from the image frame sequence, and divide each image frame into backlight areas according to the brightness highlight channel; The area detection unit is used to calculate the backlight motion mask for the backlight area to obtain the backlight motion area; The backlight adjustment unit is used to adjust the backlight motion area.

[0018] The beneficial effects of the present invention are as follows: the present invention provides a method and system for dimming an LED backlight source, obtains an image frame sequence for LED screen display, extracts a brightness highlight channel from the image frame sequence, and divides each image frame into backlight areas according to the brightness highlight channel; calculates a backlight motion mask for the backlight area, obtains a backlight motion area, and adjusts the backlight motion area. While adjusting the brightness, the hue and color saturation are considered to improve the color reproduction, especially in low-brightness scenes, to enhance the image details. Through precise image segmentation and backlight area matching, brightness leakage is reduced and visual comfort is improved. An efficient motion detection algorithm is used to adjust the backlight brightness in real time to reduce smearing and motion blur. A simple and efficient algorithm is used to adapt to the performance limitations of embedded hardware and achieve real-time processing. A smooth transition is achieved between backlight areas to avoid visual discomfort caused by discontinuous brightness. Based on this, it is suitable for real-time operation on embedded devices, which significantly improves image quality and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and other features of the present invention will become more obvious by describing in detail the embodiments shown in the accompanying drawings. The same reference numerals in the accompanying drawings of the present invention represent the same or similar elements. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work. In the accompanying drawings: Figure 1 Shown is a flow chart of a method for dimming an LED backlight source; Figure 2 Shown is a system structure diagram of an LED backlight dimming system. DETAILED DESCRIPTION

[0020] The following will be combined with the embodiments and drawings to clearly and completely describe the concept, specific structure and technical effects of the present invention, so as to fully understand the purpose, scheme and effect of the present invention. It should be noted that the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0021] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0022] like Figure 1 The flowchart of a method for dimming an LED backlight source according to the present invention is shown below. Figure 1 A method and system for dimming an LED backlight source according to an embodiment of the present invention will be described.

[0023] The present invention provides a method for dimming an LED backlight source, and the method specifically comprises the following steps: Acquire an image frame sequence for LED screen display, extract a brightness highlight channel from the image frame sequence, and divide each image frame into backlight areas according to the brightness highlight channel; Calculate the backlight motion mask for the backlight area to obtain the backlight motion area; Adjust the backlight motion zone.

[0024] Furthermore, the image output after the backlight motion area is adjusted is processed to minimize the halo effect.

[0025] In some embodiments, for images taken at night with low contrast, gradient adaptive convolution can be used for edge enhancement and bilateral kernel can be used for texture enhancement, for example, including but not limited to using APSF-guided halo rendering method for processing. The diffraction model of light (Airy point spread function) is used to accurately simulate the halo effect, so that unnecessary halo can be effectively reduced during processing.

[0026] Furthermore, the image frame sequence used for LED screen display is a time series of an image matrix, which is composed of pixel values ​​of each row and column. Brightness, hue and color saturation are separated from the pixel values, and a brightness highlighting channel is obtained by combining the contrast of brightness with hue and color saturation.

[0027] In some embodiments, the image matrix is ​​divided into R, G and B channels by the pixel values ​​of each row and column, and the value of the RGB channel of each pixel is converted into the YUV color space, with the value of Y representing brightness, the value of U representing hue, and the value of V representing color saturation. In traditional technology, only the brightness value is generally used to adjust the LED backlight. Although the visual system is much less sensitive to chromaticity than brightness, ignoring hue and color saturation will result in the backlight brightness being very accurate but the color being distorted, especially when the screen is to display some night-time pictures, it is often difficult to recognize.

[0028] Therefore, in some preferred embodiments, for each pixel point, the brightness feature can be extracted by a nonlinear function using the brightness value, the hue feature can be extracted by a nonlinear function using the hue value, and the color saturation feature can be extracted by a nonlinear function using the color saturation value; The ratio of the brightness feature to the hue feature is used as the hue backlight ratio, and the ratio of the brightness feature to the color saturation feature is used as the color saturation backlight ratio; The average of the hue backlight ratio and the color saturation backlight ratio corresponding to each pixel is used as the brightness highlight channel corresponding to each pixel.

[0029] Calculating the brightness highlight channel corresponding to each pixel in this way has another advantage: the numerical value used to represent the brightness can be effectively compressed. Since the original numerical values ​​in both RGB and YUV formats are relatively large, converting them into a combination of hue backlight ratio and color saturation backlight ratio is beneficial for representing the prominent features of brightness contrast to hue and the prominent features of brightness contrast to color saturation in smaller numerical features.

[0030] Compared with the method of controlling the backlight only by brightness, it adds considerations of hue and color saturation to avoid color distortion, especially in night scenes, objects are easier to identify. By calculating the backlight ratio, the originally large numerical range is compressed, the efficiency of calculation and storage is improved, and it is suitable for real-time processing needs. The human eye has perception of both brightness and chromaticity. Combining hue and saturation can provide a more natural visual experience. The human eye's perception of brightness, hue, and saturation is not linear. Using nonlinear functions to extract features can more accurately simulate human eye perception. The YUV color space separates brightness and chromaticity for separate processing. In this regard, the ratio of brightness to chromaticity features emphasizes the prominence of brightness in the overall color, which helps to balance brightness and color in backlight control.

[0031] Furthermore, the method for dividing each image frame into backlight areas according to the brightness highlight channel is: using an image segmentation method based on wavelet analysis and wavelet transform to perform image segmentation and obtain a number of backlight areas.

[0032] In some embodiments, the values ​​of the brightness highlight channels corresponding to each pixel point on the image matrix of each image frame are aligned with the corresponding pixel point, and the image matrix is ​​segmented according to the values ​​of the brightness highlight channels corresponding to each pixel point on the image matrix, and the sub-areas obtained by segmentation are used as backlight areas. In particular, image segmentation can be performed using image segmentation methods based on wavelet analysis and wavelet transform.

[0033] Although the use of image segmentation algorithms based on deep learning can make the segmented edges more detailed and precise, there is often a risk of over-fitting, and the parameters based on deep neural networks are too large, which is not suitable for the fast transformation processing requirements of LED devices. The use of image segmentation methods based on wavelet analysis and wavelet transform can reduce the processing time between frames, without the need for large model parameters, and reduce memory and storage requirements, which is suitable for embedded LED devices. Wavelet transform has multi-resolution analysis capabilities and can simultaneously extract spatial and frequency information of images, which is suitable for edge and texture detection. Using the characteristics of wavelet coefficients, regional features in images can be effectively identified and accurate segmentation can be performed.

[0034] Furthermore, the backlight motion mask is calculated for the backlight area, and the method for obtaining the backlight motion area is specifically as follows: For the backlight area, the absolute value of the difference between the motion frame path and the motion backlight path is calculated as the backlight motion mask of the backlight area, and whether the backlight area is a backlight motion area is determined by the backlight motion mask.

[0035] In some embodiments, according to the prior art, the backlight motion mask can represent the probability mask of the backlight motion of the frame in the image frame sequence, and a self-supervised learning method based on a deep neural network, such as Masked Modeling, can be used to convolve the time series of the image matrix in the time dimension, and extract features by sliding window style on the image segmentation area through the convolution kernel, and use the extracted features as the backlight motion mask, and then cluster the extracted features to cluster the backlight motion area. However, to perform the above processing on an offline LED backlight panel, a pre-trained neural network model is required, and storing the parameters of the pre-trained model requires a large memory. These memory devices will inevitably occupy volume on the LED device and affect the appearance structure. The convolution operation of each sub-area is performed between each frame, which has a large time complexity, a high processing time cost, and the video will be stuck.

[0036] To this end, in some preferred embodiments, in the time series of image frames, the data of the currently displayed frame is obtained as Frame_t, and the data of the previous frame is Frame_{t-1}, where t is the sequence number of the image frames in the time series from small to large in chronological order, and the previous moment of moment t is t-1.

[0037] For the current moment, the numerical matrix composed of the brightness highlight channel values ​​of each pixel point on the corresponding Frame_t is obtained as its brightness highlight channel matrix Rene_t, and the brightness highlight channel values ​​of each pixel point on the corresponding Frame_{t-1} at the previous moment are obtained to form its brightness highlight channel matrix Rene_{t-1}; The eigenvector of the brightness highlight channel matrix Rene_t corresponding to time t is calculated as oul_t, the eigenvector of the brightness highlight channel matrix Rene_{t-1} corresponding to time t is calculated as oul_{t-1}, and the cosine similarity between oul_t and oul_{t-1} is calculated as the motion frame path of Frame_t relative to Frame_{t-1}. By comparing the unit eigenvectors of the image frames in the time dimension, it is helpful to quickly anchor the contours between the image frames, so as to use the motion frame path to calculate the probability of motion between the two frames.

[0038] The brightness salient channel matrix is ​​a matrix extracted from the image frame, which reflects the prominence of the brightness of each pixel relative to the hue and color saturation. By comparing the brightness salient channel matrices of two consecutive frames, the changes in the image in the time dimension, that is, the motion information, can be captured. In a video, motion is usually manifested as changes in pixel brightness between frames. By extracting the brightness salient channel, these changes can be more effectively identified.

[0039] The eigenvector is a vector obtained by reducing the dimension of the matrix, which represents the main information of the matrix. The purpose is to convert the high-dimensional brightness salient channel matrix into a low-dimensional eigenvector for efficient comparison. The eigenvector can summarize the brightness distribution characteristics of the entire frame, making it easier to calculate the similarity between frames and then determine the overall motion trend.

[0040] The motion frame path can reflect the overall change degree between the current frame and the previous frame through cosine similarity. If the similarity of the feature vectors of the two frames is high, it means that the overall change is not large; if the similarity is low, it means that the overall change has been obvious, that is, there is motion. Compared with pixel-by-pixel comparison, calculating the cosine similarity of feature vectors is more concise and suitable for real-time processing. Feature vectors represent global information, and calculating their similarity can capture the overall motion trend.

[0041] Assume that the number of backlight areas segmented on the currently displayed Frame_t is n, where the serial number of each backlight area is i, obtain the geometric center p(i) of the backlight area with serial number i, and obtain the brightness highlight channel value array calh_t corresponding to the pixel points on a diagonal line line_t in Frame_t, then obtain the straight line line(i) passing through the point p(i) and parallel to the diagonal line of Frame_t, and obtain the brightness highlight channel value array calh_(i) corresponding to the pixel points falling on line(i) in the backlight area with serial number i.

[0042] The geometric center of the backlight area can be used to determine the position of each backlight area, which is convenient for positioning in the image. When selecting the diagonal line line_t, the diagonal line usually runs through the entire image and can reflect the global brightness changes. The straight line line(i) parallel to the diagonal line can be paralleled through the geometric center to obtain representative brightness data in the backlight area. Moreover, translation or projection can align line(i) with line_t to ensure that the positions correspond during comparison. In this way, by comparing the brightness changes in the backlight area with the global brightness changes, it is determined whether there is independent movement in the area. Special changes in local areas, that is, independent movement of the backlight area, can be identified. Only the pixels on the selected line need to be calculated, which reduces the amount of data.

[0043] Project line(i) onto line_t to obtain line_t(i) composed of a number of pixels equal to that of line(i), and obtain the brightness highlight channel values ​​corresponding to each pixel on line_t(i) to form an array calh_t(i).

[0044] Preferably, line(i) can be translated onto line_t to obtain line_t(i) composed of the same number of pixels as line(i).

[0045] Then, the cosine similarity between calh_(i) and calh_t(i) is calculated as the motion backlight path of the backlight area with sequence number i in Frame_t relative to Frame_{t-1}.

[0046] Between the backlight area with serial number i in Frame_t and Frame_{t-1}, the absolute value of the difference between the motion frame path and the motion backlight path is used as the backlight motion mask of the backlight area. If the backlight motion mask of the backlight area is not less than 1, the backlight area with serial number i in Frame_t is judged to be the backlight motion area.

[0047] The back motion mask can indicate the degree of difference between the brightness change of the backlight area and the overall brightness change. By judging whether the back motion mask is not less than 1: if it is, the backlight area is considered to be a backlight motion area; if not, the backlight area is considered to be a static area. Here, the absolute value of the cosine similarity in the positive and negative intervals is 1. Once it is greater than 1, it indicates that a crossover in the linear probability direction has occurred, so that the area with significant changes can be identified as a motion area. If the brightness change of the backlight area is significantly different from the overall change, it means that the area has independent motion and requires special processing. It helps to optimize the motion area and improve the display effect. The difference calculation and threshold judgment are very simple and suitable for real-time processing.

[0048] Special adjustments are made to the motion area to reduce smearing and blurring and improve clarity. It can also reduce computing and storage requirements and is suitable for resource-constrained devices. This provides better visual effects through precise backlight adjustment.

[0049] Then, the backlight motion area is adjusted, which may include: In some prior art embodiments, the backlight motion area may be adjusted by using an image histogram equalization algorithm, for example, by using a function in OpenCV, but the effect is relatively rough.

[0050] However, in some preferred embodiments: One of them is preferably that the brightness value of each pixel in the backlight motion area can be adjusted by dividing the brightness value of the backlight motion mask of the backlight area.

[0051] Second, preferably, for the backlight motion area, the backlight motion sub-mask of each pixel is calculated, and then the brightness value of each pixel is divided by the corresponding backlight motion sub-mask to adjust the brightness, specifically: If the backlight area with serial number i in Frame_t is determined to be a backlight motion area, the values ​​of the brightness highlight channel corresponding to each pixel point in the backlight motion area are divided by the motion frame path of Frame_t relative to Frame_{t-1}, and the obtained quotients are used as the backlight motion sub-masks corresponding to each pixel point in the backlight motion area; The brightness value of each pixel in the backlight motion area is adjusted by dividing the brightness value of each pixel by the backlight motion sub-mask corresponding to each pixel.

[0052] Compared with the traditional histogram equalization method, the use of back-motion mask and sub-mask can make more precise adjustments to the brightness and improve the image quality. Adjusting the brightness according to the degree of motion can better adapt to rapidly changing scenes and reduce motion artifacts and smearing. It avoids visual discomfort caused by excessive brightness adjustment and provides smooth brightness transition. In video processing, moving areas require special processing to avoid smearing and blurring. The back-motion mask can be used to quantitatively evaluate the degree of motion. By dividing by the back-motion mask, the brightness is actually normalized so that the brightness adjustment is inversely proportional to the degree of motion to achieve balance. Using a mask to adjust the signal amplitude is a way of LED signal processing, which can effectively control the dynamic range of the signal.

[0053] Furthermore, the method for adjusting the backlight motion area is: The brightness value of each pixel in the backlight motion area is adjusted by dividing the brightness value of the backlight motion mask of the backlight area.

[0054] Furthermore, the method for adjusting the backlight motion area may be: If a backlight area in the image frame currently displayed on the screen is determined to be a backlight motion area, the values ​​of the brightness highlight channel corresponding to each pixel point in the backlight motion area are respectively divided by the motion frame path of the image frame currently displayed on the screen relative to the previous image frame, and the obtained quotients are respectively used as the backlight motion sub-masks corresponding to each pixel point in the backlight motion area; The brightness value of each pixel in the backlight motion area is adjusted by dividing it by the backlight motion sub-mask corresponding to each pixel. The method can be performed when the previous frame is being played and the next frame is calculated, or when the current frame is being played and the previous frame is calculated and adjusted.

[0055] In order to verify the actual parameters of the technical solution of the LED backlight dimming method provided by the present invention compared with the traditional method in terms of image quality, processing efficiency, user experience, etc., a series of experiments have been designed and implemented. These experiments are aimed at comparing the differences between the method of the present invention and the traditional method in different display scenarios, focusing on evaluating the experimental data in terms of image quality, processing efficiency, user experience, etc.

[0056] In the control group (traditional method), the backlight adjustment was based on the brightness channel, without considering the hue and color saturation. In the image segmentation method, a fixed grid division was used to divide the screen into backlight areas of equal size, and no dynamic segmentation was performed. In the motion detection method, a simple frame difference method was used to compare the brightness difference between the current frame and the previous frame pixel by pixel, without considering the regional characteristics. In the halo effect processing, no special processing was performed, which may lead to obvious halo effect.

[0057] In the experimental group (this method), in terms of backlight adjustment, based on the brightness highlight channel, the ratio of brightness, hue and color saturation is combined to perform refined backlight adjustment; in terms of image segmentation method, an image segmentation method based on wavelet analysis and wavelet transform is used to dynamically divide the backlight area; in terms of motion detection method, a backlight motion mask method based on feature vector and cosine similarity calculation is used to accurately identify the backlight motion area; in terms of halo effect processing, gradient adaptive convolution and bilateral filtering are used to minimize the halo effect.

[0058] In the experimental test of static high-contrast image display, colorful and high-contrast static images are selected, such as landscape photos, works of art, etc. The images are displayed on the same display device using the control group and experimental group methods respectively.

[0059] In the detection of color reproduction, a professional colorimeter is used to measure the color difference (ΔE value) of a specific color area on the screen. In the detection of brightness uniformity, the brightness values ​​at different positions of the screen are measured and the brightness uniformity index is calculated. In the subjective evaluation test, 10 observers are invited to make subjective scores (1-10 points).

[0060] In the analysis of the experimental results, for color reproduction (ΔE value), the average ΔE value of the control group was 5.6; the average ΔE value of the experimental group was reduced to 2.1, and the color accuracy was improved by about 62%. For brightness uniformity, the brightness uniformity of the control group was 85%; the experimental group increased to 92%, and the brightness distribution was more uniform. For the average subjective evaluation score, the control group was 6.5 points; the experimental group increased to 8.9 points, and the observers were more satisfied with the image quality.

[0061] In the experimental test of low-contrast image display at night, low-contrast images such as city night scenes and starry sky taken at night were selected. The methods of the control group and the experimental group were used to measure the brightness of the brightest and darkest areas and calculate the contrast. The modulation transfer function (MTF50 value) of the image was calculated to reflect the clarity of details.

[0062] The experimental results show that in terms of contrast ratio, the control group is 800:1, and the experimental group is increased to 1200:1, a 50% increase in contrast. In terms of detail clarity (MTF50 value), the control group is 0.35, and the experimental group is increased to 0.48, a 37% increase in detail clarity.

[0063] Through the above experiments, we can draw the following conclusions: This method has obvious advantages in color reproduction, the ΔE value is reduced by about 62%, the color is more realistic, the contrast is increased by 50%, and the detail clarity is improved by 37%, especially in dark scenes.

[0064] The present invention also provides some application implementation modes, including: Implementation method 1 provides an application in smart TVs, specifically: embed the backlight adjustment algorithm module of the present invention in the image processing chip or software system of the TV; the TV receives input signals, extracts brightness highlight channels in real time, and analyzes the brightness, hue and color saturation characteristics of the image; divides the backlight area according to the analysis results, calculates the backlight motion mask, and determines the backlight motion area; adjusts the backlight motion area and the static area respectively to optimize the backlight brightness and contrast; uses gradient adaptive convolution and bilateral filtering to minimize the halo effect and improve image quality; provides a user interface to allow users to select preset image modes (such as cinema mode, game mode) or manually adjust parameters. The dynamic contrast and color performance of the image are improved, and motion blur and smear are reduced, which is particularly suitable for watching sports events and action movies.

[0065] Implementation method 2 provides an application in a smart phone, specifically: using the ambient light sensor of the phone to detect the current lighting conditions. In strong light, increase the backlight brightness; in weak light, reduce the backlight brightness to protect the eyes. Real-time analysis of screen display content, extract brightness highlight channels, and dynamically adjust the backlight area. When the battery is low, reduce the backlight brightness to improve energy efficiency. Through intelligent adjustment, extend battery life.

[0066] Implementation method 3 provides an application in an in-vehicle display system, specifically: combining light sensors inside and outside the vehicle to determine the current ambient brightness. In strong daylight, increase the backlight brightness; at night or in a tunnel, reduce the backlight brightness to avoid glare. For important warning information, enhance the brightness and contrast of the backlight area to ensure that the driver notices it. Use halo effect minimization processing to reduce the visual interference of the screen to the driver. Adaptive adjustment according to the environment to provide a comfortable visual experience.

[0067] The LED backlight dimming system runs in any computing device such as a desktop computer, a laptop computer, a mobile phone, a PDA or a cloud data center. The computing device includes: a processor, a memory and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps in the LED backlight dimming method are implemented. The executable system may include, but is not limited to, a processor, a memory, and a server cluster.

[0068] An embodiment of the present invention provides a system for dimming an LED backlight source, such as Figure 2As shown, a system for dimming an LED backlight source in this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned method for dimming an LED backlight source are implemented to optimize the brightness of the LED screen. The processor executes the computer program to run in the following system units: A data extraction unit, used to obtain an image frame sequence for LED screen display, extract a brightness highlight channel from the image frame sequence, and divide each image frame into backlight areas according to the brightness highlight channel; The area detection unit is used to calculate the backlight motion mask for the backlight area to obtain the backlight motion area; The backlight adjustment unit is used to adjust the backlight motion area.

[0069] Among them, preferably, all undefined variables in the present invention, if not clearly defined, can be manually set thresholds.

[0070] Among them, dimensionless numerical calculations are used between physical quantities of different units.

[0071] The LED backlight dimming system can be run in computing devices such as desktop computers, laptops, mobile phones, PDAs, and cloud data centers. The LED backlight dimming system includes, but is not limited to, a processor and a memory. Those skilled in the art can understand that the example is only an example of a method and system for dimming an LED backlight, and does not constitute a limitation on a method and system for dimming an LED backlight. It can include more or fewer components than the example, or a combination of certain components, or different components. For example, the LED backlight dimming system can also include input and output devices, network access devices, buses, etc.

[0072] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete component gate circuits or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the LED backlight dimming system, and uses various interfaces and lines to connect the various sub-areas of the entire LED backlight dimming system.

[0073] The memory can be used to store the computer program and / or module, and the processor implements the various functions of the LED backlight dimming method and system by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0074] The present invention provides a method and system for dimming an LED backlight source, which obtains an image frame sequence for LED screen display, extracts a brightness highlight channel from the image frame sequence, divides each image frame into backlight areas according to the brightness highlight channel, calculates a backlight motion mask for the backlight area, obtains a backlight motion area, and adjusts the backlight motion area. The method is suitable for real-time operation on embedded devices, which significantly improves image quality and user experience.

[0075] Although the description of the present invention has been quite detailed and has been described in particular with respect to several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, so as to effectively cover the intended scope of the present invention. In addition, the present invention is described above with the embodiments foreseeable by the inventors, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that are not currently foreseen may still represent equivalent changes of the present invention.

Claims

1. A method for dimming an LED backlight source, characterized in that: The method comprises the following steps: Obtain an image frame sequence for LED screen display, extract a brightness highlight channel from the image frame sequence, and divide each image frame into a backlight area according to the brightness highlight channel; calculate a backlight motion mask for the backlight area, obtain a backlight motion area, and adjust the backlight motion area; The backlight motion mask represents a probability mask of backlight motion of frames in the image frame sequence.

2. The method for dimming an LED backlight source according to claim 1, characterized in that: The image output after adjusting the backlight motion area is processed to minimize the halo effect.

3. The method for dimming an LED backlight source according to claim 1, characterized in that: The method for dividing each image frame into backlight areas according to the brightness highlight channel is: using an image segmentation method based on wavelet analysis and wavelet transform to perform image segmentation and obtain a number of backlight areas.

4. The method for dimming an LED backlight source according to claim 1, characterized in that: in, The image frame sequence used for LED screen display is a time sequence of an image matrix, which is composed of pixel values ​​of each row and column. Brightness, hue and color saturation are separated from the pixel values, and the brightness highlight channel is obtained by combining the contrast of brightness with hue and color saturation.

5. The method for dimming an LED backlight source according to claim 1 or 4, characterized in that: The method of calculating the backlight motion mask for the backlight area and obtaining the backlight motion area is as follows: For the backlight area, the absolute value of the difference between the motion frame path and the motion backlight path is calculated as the backlight motion mask of the backlight area, and whether the backlight area is a backlight motion area is determined by the backlight motion mask; Among them, the motion frame path can be used to represent the overall motion degree between frames based on the brightness highlight channel matrix of the current frame and the previous frame, and the motion backlight path can be used to represent the motion degree of the backlight area relative to the global based on the brightness highlight channel data in the backlight area and the brightness data at the corresponding position on the global diagonal.

6. The method for dimming an LED backlight source according to claim 5, characterized in that: The method to adjust the backlight motion area is: The brightness value of each pixel in the backlight motion area is adjusted by dividing the brightness value of the backlight motion mask of the backlight area.

7. The method for dimming an LED backlight source according to claim 5, characterized in that: The method for adjusting the backlight motion area is either: If there is a backlight area in the image frame currently displayed on the screen and it is determined to be a backlight motion area, the values ​​of the brightness highlight channel corresponding to each pixel point in the backlight motion area are respectively divided by the motion frame path of the image frame currently displayed on the screen relative to the previous image frame, and the obtained quotients are respectively used as the backlight motion sub-masks corresponding to each pixel point in the backlight motion area; The brightness value of each pixel in the backlight motion area is adjusted by dividing the brightness value of each pixel by the backlight motion sub-mask corresponding to each pixel.

8. A system for dimming an LED backlight source, characterized in that: The LED backlight dimming system runs in any computing device such as a desktop computer, a laptop computer or a cloud data center. The computing device includes: a processor, a memory and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps in the LED backlight dimming method as described in any one of claims 1 to 4 are implemented.

Images