Method, system and chip for reducing power consumption of LCD panel timing control chip

Abstract

The present invention relates to the field of liquid crystal display technology, and discloses a method, system, and chip for reducing the power consumption of a timing control chip of a liquid crystal panel, including obtaining historical viewing data of a user, building and training a user visual preference prediction model; collecting the user's visual attention data and display content data in real time, and determining the focus area on the liquid crystal panel in combination with the prediction model; dividing the liquid crystal panel into high, medium, and low resolution areas and setting corresponding pixel density and refresh frequency thresholds; performing visual saliency analysis on the display content to determine key salient areas and non-key salient areas; performing local refresh in different resolution areas, wherein the refresh frequency of the key salient area is higher than the refresh frequency of the non-key salient area. The system includes multiple modules to implement the above method. The chip adopts the system and method, and can effectively reduce the power consumption of the timing control chip of the liquid crystal panel.

Description

Technical Field

[0001] The present invention relates to the field of liquid crystal display technology, and more specifically, to a method, a system, and a chip for reducing power consumption of a timing control chip of a liquid crystal panel. Background Art

[0002] With the continuous development of liquid crystal display technology, people have put forward higher requirements on the power consumption and user experience of display devices. How to reduce the power consumption of LCD panel timing control chips while providing better display effects and user experience has become a hot topic in current research.

[0003] At present, some technologies have been tried to solve the above problems. For example, patent IN201747023657A discloses an efficient operation method of a wearable display, which displays images with different image qualities in different display areas of the wearable display device to enhance user experience and extend battery life. In addition, the Chinese patent application with publication number CN114399973A discloses an adaptive display screen based on environmental perception, which adjusts the operation of the display screen through environmental perception and user perception, and then performs data fusion based on the perceived data to realize intelligent adaptive adjustment of the display screen.

[0004] Although the existing technology has optimized the display effect and power consumption to a certain extent, there are still some shortcomings. The existing LCD panel timing control chip usually adopts a fixed driving method, fails to fully consider the user's visual preferences and attention distribution, and does not perform differentiated processing based on the importance of the displayed content, resulting in high power consumption and poor user experience. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method, a system and a chip for reducing the power consumption of a timing control chip of a liquid crystal panel.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for reducing power consumption of a timing control chip of a liquid crystal panel comprises:

[0008] Acquire the user's first historical viewing data, build and train a user visual preference prediction model; collect the user's visual attention data and display content data in real time, and determine the first focus area ROI1 and the second focus area ROI2 on the liquid crystal panel in combination with the user's visual preference prediction model;

[0009] Divide the liquid crystal panel into high-resolution area, medium-resolution area and low-resolution area, set corresponding pixel density thresholds and refresh frequency thresholds for different resolution areas, and dynamically adjust the pixel density and refresh frequency in each resolution area;

[0010] Visual saliency analysis is performed on the display content of the liquid crystal panel to determine key salient areas and non-key salient areas; local refresh is performed in different resolution areas, wherein the refresh frequency of the key salient areas is higher than the refresh frequency of the non-key salient areas.

[0011] Further, the determining of the first region of interest ROI1 and the second region of interest ROI2 on the liquid crystal panel includes:

[0012] Input the real-time collected visual attention data and display content data into the user visual preference prediction model to obtain the predicted visual preference distribution map and attention area coordinates;

[0013] According to the predicted visual preference distribution map and the coordinates of the regions of interest, a first region of interest ROI1 and a second region of interest ROI2 on the liquid crystal panel are determined.

[0014] Furthermore, the method for dividing the liquid crystal panel into high-resolution area, medium-resolution area and low-resolution area is: dividing the area where ROI1 is located into a high-resolution area, dividing the area where ROI2 is located into a medium-resolution area, and dividing the non-ROI area into a low-resolution area.

[0015] Furthermore, the region where ROI1 is located is divided into a high-resolution region by taking the center of ROI1 as the center of the circle and the size of ROI1 as the radius to divide a circular or elliptical high-resolution region; the high-resolution region completely covers ROI1;

[0016] The region where ROI2 is located is divided into a medium-resolution region by taking the center of ROI2 as the center of the circle and the size of ROI2 as the radius to divide a circular or elliptical medium-resolution region; the medium-resolution region completely covers ROI2.

[0017] Furthermore, the setting of corresponding pixel density thresholds and refresh frequency thresholds for different resolution areas includes:

[0018] Set the pixel density lower limit of the high-resolution area to PPI H , the refresh rate is limited to FPS H ;

[0019] Set the pixel density lower limit of the medium resolution area to PPI M , capped at PPI H , the refresh rate is limited to FPS M ;

[0020] Set the pixel density limit of low-resolution areas to PPI L , the refresh rate is capped at FPS L ;

[0021] Among them, PPI L <PPI AV <PPI M <PPI H , PPI AV is the average pixel density of the LCD panel, FPS L <FPS M <FPS H .

[0022] Furthermore, the dynamically adjusting the pixel density and refresh rate in each resolution area includes:

[0023] Adaptively adjust the pixel density P in the high-resolution area H , satisfying P H ≥PPI H ; Adaptive adjustment of refresh rate F H , satisfying F H ≥FPS H ;

[0024] Adaptively adjust the pixel density P in the medium resolution area M , meet PPI M ≤P M <PPI H ; Adaptive adjustment of refresh rate F M , meet FPS M ≤F M <FPS H ;

[0025] Adaptively adjust the pixel density P in the low-resolution area L , satisfying P L <PPI L ; Adaptive adjustment of refresh rate F L , satisfying F L <FPS L ;

[0026] The pixel density P is adaptively adjusted in the high-resolution area. H and refresh rate F H The methods include:

[0027]

[0028]

[0029] Among them, A ROI1 is the area of ​​the first region of interest ROI1, A Total is the total display area of ​​the LCD panel, T ROI1 is the user's gaze time on the first region of interest ROI1, T Totalis the total gaze time of the user on the entire LCD panel, α1 is the pixel density sensitivity coefficient, and β1 is the refresh rate sensitivity coefficient.

[0030] Furthermore, the determining of the key salient area and the non-key salient area includes:

[0031] Extract multi-scale visual features of the content displayed on the LCD panel;

[0032] Based on the extracted multi-scale visual features, the local saliency map and the global saliency map of the displayed content are calculated;

[0033] Generate the final visual saliency map based on the local saliency map and the global saliency map;

[0034] A significance threshold is set. The areas in the visual significance map whose significance is greater than the significance threshold are marked as key significant areas, and the areas whose significance is less than or equal to the significance threshold are marked as non-key significant areas.

[0035] Further, the performing local refreshing in different resolution areas includes:

[0036] In the high-resolution area, the first high frequency F is used for the key salient area. H1 Perform local refresh and use the first low frequency F for non-critical salient areas L1 Perform a partial refresh; among them, FPS H ≤F L1 <F H1 ;

[0037] In the medium resolution area, the second highest frequency F is used for the key salient area. H2 Perform local refresh and use the second lowest frequency F for non-critical salient areas L2 Perform a partial refresh; among them, FPS M ≤F L2 <F H2 <FPS H ;

[0038] In the low-resolution area, the third lowest frequency F is used for the key salient area. L3 Perform local refresh and use the fourth lowest frequency F for non-critical significant areas L4 Perform a partial refresh; where F L4 <F L3 <FPS L .

[0039] A system for reducing the power consumption of a timing control chip of a liquid crystal panel, which is used to implement the above-mentioned method for reducing the power consumption of a timing control chip of a liquid crystal panel, and the system comprises:

[0040] Real-time data collection module: used to collect user's visual attention data and display content data in real time;

[0041] Visual preference prediction module: used to obtain the user's first historical viewing data, build and train the user's visual preference prediction model; determine the first focus area ROI1 and the second focus area ROI2 on the LCD panel according to the visual attention data and the display content data, combined with the user's visual preference prediction model;

[0042] Resolution area division module: used to divide the LCD panel into high-resolution area, medium-resolution area and low-resolution area;

[0043] Dynamic adjustment module: used to set corresponding pixel density thresholds and refresh rate thresholds for different resolution areas, and dynamically adjust the pixel density and refresh rate in each resolution area;

[0044] Visual saliency analysis module: used to perform visual saliency analysis on the display content of the LCD panel and determine the key saliency area and the non-key saliency area;

[0045] Local refresh module: used to perform local refresh in different resolution areas, where the refresh frequency of key salient areas is higher than that of non-key salient areas.

[0046] A chip comprises the above-mentioned system for reducing the power consumption of a liquid crystal panel timing control chip.

[0047] An electronic device comprises the above-mentioned chip, memory, central processing unit and a computer program stored in the memory and executable on the central processing unit. When the central processing unit executes the computer program, the above-mentioned method for reducing the power consumption of the timing control chip of the liquid crystal panel is implemented.

[0048] A computer-readable storage medium stores a computer program, and when the computer program is executed, the method for reducing the power consumption of a timing control chip of a liquid crystal panel is implemented.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention significantly reduces the power consumption of the timing control chip of the liquid crystal panel by dynamically adjusting the pixel density and refresh frequency of different areas of the liquid crystal panel; in the key areas (ROI1 and ROI2) that users pay attention to, the resolution and refresh frequency are improved to ensure image quality; in non-key areas, the resolution and refresh frequency are reduced to reduce energy consumption; compared with the traditional constant refresh frequency, the present invention maximizes the reduction of power consumption while ensuring user experience.

[0051] The present invention collects the user's visual attention data in real time, combines it with the user's visual preference prediction model, and dynamically adjusts the display effect of the liquid crystal panel, so that the user can obtain the best visual experience when watching; the high-resolution area provides higher image clarity and smoothness, while the low-resolution area reduces unnecessary power consumption; this intelligent adjustment method not only improves the user's viewing experience, but also extends the use time of the device.

[0052] The present invention adopts an intelligent user visual preference prediction model and a visual saliency analysis module, which can perform adaptive adjustments according to the user's actual usage. By analyzing the user's historical viewing data and real-time visual attention data, the system can predict the areas that the user is most likely to pay attention to and optimize the display of these areas. This adaptive adjustment method enables the liquid crystal panel to intelligently adapt to the visual needs of different users and provide personalized display effects.

[0053] Since the present invention can effectively reduce the power consumption of the LCD panel timing control chip, the heat generation of the chip and panel is reduced, and equipment aging and failure caused by high temperature are avoided; therefore, the present invention not only improves the energy efficiency of the LCD panel, but also extends the overall service life of the equipment and reduces maintenance and replacement costs.

[0054] The present invention can accurately identify key salient areas and non-key salient areas in an image by performing visual saliency analysis on displayed content; a higher refresh rate and resolution are used in the key salient areas to ensure image clarity and smoothness, thereby improving the overall viewing effect; meanwhile, a lower refresh rate and resolution are used in the non-key salient areas, which can still meet basic visual needs but greatly reduce power consumption.

[0055] The method and system of the present invention are not only applicable to traditional liquid crystal panels, but can also be flexibly applied to various new display devices, such as wearable devices, car displays, smart home display screens, etc. Different application scenarios have different requirements for display effects and power consumption. The adaptive adjustment mechanism of the present invention can be optimized according to specific needs to provide a flexible and efficient solution.

[0056] The present invention introduces a user visual preference prediction model and visual saliency analysis technology, combined with real-time data acquisition and dynamic adjustment modules, to achieve intelligent control of the display effect of the liquid crystal panel; this technical means is not only innovative in power consumption management, but also provides new ideas and methods for the future development of display technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 It is a flow chart of the method for reducing the power consumption of the timing control chip of the liquid crystal panel in the present invention;

[0059] Figure 2 A flow chart of a method for dynamically adjusting the pixel density and refresh rate in each resolution area in a method for reducing the power consumption of a timing control chip of a liquid crystal panel according to the present invention;

[0060] Figure 3 A flow chart of a method for performing visual significance analysis on display content of a liquid crystal panel in a method for reducing power consumption of a timing control chip of a liquid crystal panel according to the present invention;

[0061] Figure 4 It is a functional module diagram of the system for reducing the power consumption of the timing control chip of the liquid crystal panel in the present invention. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Example 1

[0064] See also Figure 1 As shown, this embodiment provides a method for reducing the power consumption of a timing control chip of a liquid crystal panel, including:

[0065] Step S1000, obtaining the user's first historical viewing data, building and training a user visual preference prediction model; collecting the user's visual attention data and display content data in real time, and determining a first focus area ROI1 and a second focus area ROI2 on the liquid crystal panel in combination with the user's visual preference prediction model;

[0066] Further, step S1000 includes:

[0067] Step S1100, obtaining the first historical viewing data of the user;

[0068] Furthermore, step S1100 includes:

[0069] Step S1110, recording the changes in the sight focus position of the user when viewing the displayed content, to form a sight focus position sequence;

[0070] Step S1220, analyzing the user's attention distribution in different areas to form an attention distribution sequence;

[0071] Step S1130, analyzing the displayed content viewed by the user, extracting its visual features, and forming a displayed content feature sequence;

[0072] Step S1140 , collecting users' subjective evaluation and preference feedback on different display contents and display qualities, and forming a visual preference feedback sequence.

[0073] Specifically, in step S1100, the visual behavior data of the user in the actual viewing scene is obtained through multimodal data collection; the sight focus position sequence reflects the spatial distribution and transfer trajectory of the user's visual attention on the display content, and the eye tracking device can be used to record the coordinates of the user's sight position in real time. Based on the sight focus, the attention distribution sequence further quantifies the user's attention allocation intensity to different areas, which can be intuitively represented by an attention heat map. The color or brightness in the heat map represents the relative intensity of attention. The display content feature sequence characterizes the multidimensional visual attributes of the display content, such as the color histogram representing the color distribution, the gradient histogram representing the texture complexity, and the motion vector representing the dynamic characteristics, etc. These characteristics are closely related to visual perception and attention. The visual preference feedback sequence records the user's subjective evaluation of different display effects, which can be obtained through explicit user scoring, questionnaires, etc., or inferred by analyzing the user's implicit interactive behavior (such as adjusting display settings, viewing time, etc.). By collecting these multi-dimensional historical viewing data, the user's visual behavior pattern and preference characteristics in the actual viewing scene can be fully characterized, providing a data basis for subsequent display optimization.

[0074] Step S1200, constructing and training a user visual preference prediction model based on a deep learning algorithm;

[0075] Specifically, in step S1200, a deep learning algorithm is used to extract patterns and regularities of user visual preferences from historical viewing data. First, the collected first historical viewing data is divided into three subsets: training set, validation set and test set, where the training set is used for model training, the validation set is used for model hyperparameter tuning and performance evaluation, and the test set is used to evaluate the generalization ability of the model. Then, according to the characteristics of the historical viewing data and the needs of the prediction task, a suitable deep learning model architecture is designed. Commonly used models include convolutional neural networks (CNN), long short-term memory networks (LSTM), attention mechanisms, etc. CNN is suitable for processing image data and can extract visual features of displayed content; LSTM is suitable for processing time series data and can model the time dependency of user visual behavior; attention mechanisms can adaptively focus on the importance of different areas, which is consistent with the user's attention allocation. When designing the model, it is necessary to comprehensively consider feature extraction, time series modeling and attention mechanisms to make full use of the information of multi-dimensional data. The designed model is trained using the training set data, and the prediction output is calculated through forward propagation. The model parameters are adjusted through back propagation so that the model can fit the relationship between the user's visual preference and the viewing data. Commonly used optimization algorithms include stochastic gradient descent (SGD) and adaptive moment estimation (Adam). During the model training process, the prediction performance of the model is evaluated regularly using the validation set data, such as calculating indicators such as accuracy, precision, and recall. According to the verification results, the model's hyperparameters (such as learning rate, regularization coefficient, etc.) are adjusted to select the model configuration with the best performance. Finally, after the training is completed, the generalization ability of the model is verified using the test set data to evaluate its prediction performance and generalization ability on new data. The test results can reflect the actual application effect of the model.

[0076] Through the training of deep learning models, the patterns and rules of user visual preferences can be automatically extracted and learned from massive amounts of historical viewing data, and a mapping relationship between input viewing data and output visual preferences can be established. The trained model can predict the user's interest in the current display content and focus area based on the user's real-time viewing behavior, providing a decision-making basis for subsequent adaptive display optimization. The powerful representation ability and automatic feature extraction capabilities of deep learning enable the model to mine implicit visual preference patterns from complex, high-dimensional viewing data and provide personalized display optimization strategies. At the same time, the deep learning model has good generalization ability and can effectively predict and adapt to new and unknown viewing scenarios based on the learning results of historical data.

[0077] Step S1300, collecting the user's visual attention data and display content data in real time, and determining a first focus area ROI1 and a second focus area ROI2 on the liquid crystal panel in combination with a user visual preference prediction model;

[0078] Furthermore, step S1300 includes:

[0079] Step S1310, recording the eye movement index of the user in real time to obtain the visual attention data of the user when viewing the displayed content;

[0080] Step S1320, capturing the current display content of the liquid crystal panel, extracting its visual features, and forming display content data;

[0081] Step S1330, inputting the real-time collected visual attention data and display content data into the user visual preference prediction model to obtain a predicted visual preference distribution map and attention area coordinates;

[0082] Step S1340 , determining a first region of interest ROI1 and a second region of interest ROI2 on the liquid crystal panel according to the predicted visual preference distribution map and the coordinates of the regions of interest.

[0083] Specifically, through eye tracking technology, such as video analysis, infrared reflection and other methods, the user's eye gaze point coordinates, gaze time, blink frequency, pupil size change and other eye movement indicators are recorded in real time. These indicators reflect the spatial distribution and temporal changes of the user's visual attention on the displayed content; at the same time, the current displayed images, videos and other contents are collected and analyzed in real time to extract their visual features, such as color histogram, brightness distribution, texture complexity, motion information, etc. These features characterize the visual attributes of the current displayed content and are important factors affecting the user's visual perception and attention allocation. The visual attention data and display content data are input into the pre-trained visual preference prediction model. The model can predict the user's interest in the current displayed content and the area that may be concerned based on the user preference pattern learned from the historical viewing data, and generate a visual preference distribution map and attention area coordinates. The visual preference distribution map is a matrix of the same size as the displayed content, in which the value of each element represents the predicted value of the attention at the corresponding position. The attention area coordinates represent the position and size of the predicted high attention area in the displayed content. According to the prediction results, the first two regions with the highest attention are selected and marked as ROI1 and ROI2 respectively. ROI1 represents the region that the user is most interested in and most likely to pay attention to, and ROI2 represents the region that the user is second most interested in and second most likely to pay attention to.

[0084] By collecting and predicting the user's visual attention in real time, it is possible to dynamically adapt to changes in the user's viewing behavior and adjust the display optimization strategy in a timely manner. The prediction model makes full use of the user's historical preferences and current viewing status, and can predict the user's real visual needs with a high degree of accuracy. Combining attention prediction with display content analysis can fully consider the visual influencing factors of the display content and the user's subjective preferences, and provide more accurate and personalized ROI recognition results. The determination of ROI1 and ROI2 provides key optimization areas for subsequent display optimization methods such as adaptive resolution control, local refresh, and backlight adjustment. It can concentrate resources to ensure the display quality of key areas, while reducing power consumption in non-key areas, and achieving a balance between image quality and energy consumption.

[0085] Step S2000, dividing the liquid crystal panel into a high-resolution area, a medium-resolution area, and a low-resolution area, setting corresponding pixel density thresholds and refresh frequency thresholds for different resolution areas, and dynamically adjusting the pixel density and refresh frequency in each resolution area;

[0086] Furthermore, step S2000 includes:

[0087] Step S2100, dividing the area where ROI1 is located into a high-resolution area, dividing the area where ROI2 is located into a medium-resolution area, and dividing the non-ROI area into a low-resolution area;

[0088] Furthermore, step S2100 includes:

[0089] Step S2110, taking the center of ROI1 as the center of the circle and the size of ROI1 as the radius, a circular or elliptical high-resolution area is divided; the high-resolution area completely covers ROI1 and a certain margin is reserved around it;

[0090] Step S2120: divide a circular or elliptical medium-resolution area with the center of ROI2 as the center and the size of ROI2 as the radius; the medium-resolution area completely covers ROI2.

[0091] Specifically, the area of ​​the high-resolution area is slightly larger than ROI1 to ensure the display quality of the area of ​​interest and its surroundings. The area of ​​the medium-resolution area is equivalent to ROI2 to ensure the display quality of the secondary area of ​​interest, while controlling the area to balance power consumption. The remaining areas except the high-resolution area and the medium-resolution area are divided into low-resolution areas. The low-resolution area covers most of the panel area. By reducing the resolution of these non-interest areas, power consumption can be significantly reduced. It should be noted that the division of high-resolution area, medium-resolution area, and low-resolution area is dynamic and needs to be adjusted in real time according to the movement of the user's line of sight and focus. Through dynamic resolution division, display resources can be concentrated on the area of ​​interest to the user, improving the visual experience, while reducing power consumption in non-interest areas and extending battery life.

[0092] Step S2200, setting corresponding pixel density thresholds and refresh rate thresholds for different resolution areas;

[0093] Further, step S2200 includes:

[0094] Step S2210, setting the pixel density lower limit of the high-resolution area to PPI H , the refresh rate is limited to FPS H ;

[0095] Step S2220, setting the lower limit of pixel density in the medium resolution area to PPI M , capped at PPI H , the refresh rate is limited to FPS M ;

[0096] Step S2230, setting the upper limit of pixel density of the low-resolution area to PPI L , the refresh rate is capped at FPS L ; Among them, PPI L <PPI AV <PPI M <PPI H , PPI AV is the average pixel density of the LCD panel, FPS L <FPS M <FPS H .

[0097] Specifically, PPI H and FPS H The specific value of needs to take into account the physical characteristics of the panel, display requirements, power consumption budget and other factors. H Can provide more delicate display details and higher FPS H It can ensure smooth playback of dynamic content, such as videos, games and other scenarios. PPIM Usually lower than PPI H , but higher than the average pixel density of the panel to ensure that the display effect of MRR is better than that of non-interest areas. FPS M Can be lower than FPS H , but a certain level of dynamic display performance must be guaranteed, such as smooth text scrolling, image browsing, etc. PPI L and FPS L Usually set low to minimize power consumption. L Basic display clarity and FPS must be guaranteed L It is necessary to ensure basic display smoothness to avoid problems such as screen freeze and tearing. Reasonable setting of the pixel density and refresh rate thresholds in the resolution area can achieve the best balance between display quality and power consumption. By adopting high pixel density and high refresh rate in the high-resolution area, medium pixel density and refresh rate in the medium-resolution area, and low pixel density and low refresh rate in the low-resolution area, the overall power consumption can be effectively reduced while ensuring the display experience in the area of ​​user attention. The specific value of the threshold needs to be optimized based on factors such as panel characteristics, user needs, and power consumption budget through a large number of subjective and objective experiments and user feedback. In addition, the threshold can be dynamically adjusted according to different display content and scenarios to further optimize the display effect and energy efficiency.

[0098] Step S2300, obtaining the pixel density and refresh rate in each resolution area, and dynamically adjusting the pixel density and refresh rate in each resolution area according to the corresponding pixel density threshold and refresh rate threshold set for different resolution areas;

[0099] Furthermore, if Figure 2 As shown, step S2300 includes:

[0100] Step S2310: adaptively adjust the pixel density P in the high-resolution area H , satisfying P H ≥PPI H ; Adaptive adjustment of refresh rate F H , satisfying F H ≥FPS H ;

[0101] Step S2320: adaptively adjust the pixel density P in the medium resolution area. M , meet PPI M ≤P M <PPI H ; Adaptive adjustment of refresh rate F M , meet FPS M ≤F M <FPS H ;

[0102] Step S2330: adaptively adjust the pixel density P in the low-resolution area. L , satisfying P L <PPI L ; Adaptive adjustment of refresh rate F L , satisfying F L <FPS L .

[0103] Specifically, in high-resolution areas, the pixel density and refresh rate are adaptively adjusted according to the complexity and dynamic characteristics of the displayed content. For content with rich details and high-frequency changes, such as text, graphics, videos, etc., the pixel density and refresh rate can be appropriately increased to show more details and motion information. However, the increase in pixel density and refresh rate needs to be weighed against the increase in power consumption and should not exceed the power consumption budget. For content with sparse details and low-frequency changes, such as backgrounds, icons, etc., the pixel density and refresh rate can be appropriately reduced to save power consumption, but the pixel density should not be lower than the PPI. H , the refresh rate should not be lower than FPS H To ensure the basic display quality of the high-resolution area. In the medium-resolution area, the pixel density and refresh rate are adjusted in real time according to the characteristics of the displayed content. Unlike the high-resolution area, the pixel density adjustment range of the medium-resolution area is within PPI. M and PPI H to strike a balance between display quality and power consumption. The refresh rate in the medium resolution area is no less than FPS M To meet certain dynamic display requirements. In the low-resolution area, since its main goal is to reduce power consumption, both the pixel density and refresh rate should be as low as possible. However, in order to ensure basic display quality and smoothness, the pixel density should not be higher than PPI. L , the refresh rate should not be higher than FPS L .

[0104] Dynamically adjusting the pixel density and refresh rate within the resolution area can further optimize the display effect and power consumption according to the characteristics of the displayed content and the real-time needs of the user. By appropriately increasing the pixel density and refresh rate in the high-resolution area, more details and dynamic information can be displayed, improving the user's viewing experience; by balancing the pixel density and refresh rate in the medium-resolution area, power consumption can be reduced while ensuring a certain display quality; by reducing the pixel density and refresh rate as much as possible in the low-resolution area, the power consumption of non-focus areas can be minimized. The size of the adjustment range needs to take into account factors such as panel capabilities, display requirements, and power consumption budget, and is determined through repeated engineering optimization and user verification. In addition, intelligent algorithms can be introduced to automatically infer the optimal pixel density and refresh rate based on factors such as the complexity of the displayed content, dynamic characteristics, and user preferences, simplify the parameter adjustment process, and improve adaptation efficiency.

[0105] Adaptively adjust the pixel density P in the high-resolution area H and refresh rate F H The methods include:

[0106]

[0107]

[0108] in:

[0109] P H : Pixel density of high-resolution area;

[0110] F H : Refresh rate of high-resolution area;

[0111] A ROI1 : The area of ​​the first region of interest ROI1;

[0112] A Total : The total display area of ​​the LCD panel;

[0113] T ROI1 : The user's gaze time on the first focus area ROI1;

[0114] T Total : The total gaze time of the user on the entire LCD panel;

[0115] α1: pixel density sensitivity coefficient, determined by technicians in this field through a large number of experiments;

[0116] β1: refresh frequency sensitivity coefficient, determined by technicians in this field through a large number of experiments;

[0117] When the user's gaze time on the first region of interest ROI1 is T ROI1As the pixel density P increases, H and refresh rate F H will increase accordingly, improving the display quality of the area; when the area of ​​the first region of interest ROI1 increases, the pixel density P H and refresh rate F H The value of the sensitivity coefficient (α1 and β1) affects the sensitivity of the formula. By adjusting these coefficients, the display quality and power consumption can be balanced. The formula dynamically adjusts the pixel density and refresh rate according to the user's real-time visual attention data to optimize the user's display experience. By reasonably adjusting the display parameters of non-focused areas, power consumption can be significantly reduced, the battery life of the device can be extended, and the display parameters can be improved in the areas where the user is concerned, ensuring high-quality display effects and improving the user's visual experience.

[0118] Step S3000, performing visual saliency analysis on the display content of the liquid crystal panel to determine key salient areas and non-key salient areas; performing local refresh in different resolution areas, wherein the refresh frequency of the key salient areas is higher than the refresh frequency of the non-key salient areas;

[0119] Furthermore, step S3000 includes:

[0120] Step S3100, performing visual saliency analysis on the display content of the liquid crystal panel to determine key salient areas and non-key salient areas;

[0121] For further information, see Figure 3 , step S3100 includes:

[0122] Step S3110, extracting multi-scale visual features of the content displayed on the liquid crystal panel;

[0123] Step S3120, calculating a local saliency map and a global saliency map of the displayed content based on the extracted multi-scale visual features;

[0124] Step S3130, generating a final visual saliency map according to the local saliency map and the global saliency map;

[0125] Step S3140, setting a significance threshold, marking the area in the visual significance map whose significance is greater than the significance threshold as a key significance area, and marking the area whose significance is less than or equal to the significance threshold as a non-key significance area.

[0126] The method for generating a final visual saliency map according to the local saliency map and the global saliency map comprises:

[0127] S(x,y)=α×L(x,y)+β×G(x,y)+γ×C(x,y)×∫∫(ω(xu,yv)×(L(x,y)-L(u,v)) 2 )dudv

[0128] in:

[0129] S(x, y): represents the final visual saliency value at the image coordinate (x, y);

[0130] L(x, y): represents the local saliency value at the image coordinate (x, y), which is calculated by the center-surround difference algorithm;

[0131] G(x, y): represents the global saliency value at the image coordinate (x, y), which is calculated by frequency domain analysis or graph theory analysis method;

[0132] C(x, y): is a normalization constant used to balance the contribution of local significance and global significance to the final result;

[0133] α: is a parameter that controls the local significance weight, which is determined by technicians in this field through a large number of experiments;

[0134] β: is a parameter that controls the global significance weight, which is determined by technicians in this field through a large number of experiments;

[0135] γ: context modulation coefficient, which controls the modulation intensity of context information on local saliency and is determined by those skilled in the art through a large number of experiments;

[0136] ω(xu, yv): is a two-dimensional Gaussian kernel function that measures the spatial distance weight between the image coordinate (x, y) and its neighborhood (u, v);

[0137] The formula fuses the local saliency map L and the global saliency map G in a linear weighted manner, taking into account the saliency of the image region at both local and global scales. At the same time, the context modulation factor ∫∫(ω(xu,yv)×(L(x,y)-L(u,v)) is introduced. 2)dudv, the context modulation factor takes into account the difference in local saliency features between the image region and its neighborhood. The greater the difference, the higher the saliency of the central region. The Gaussian kernel function ω(x, y) decays with the increase of spatial distance, so that the neighboring region has a greater influence on the central region. The normalization constant C(x, y) is used to balance the local saliency L(x, y) and the context modulation factor to adapt to the dynamic range of the image. As the local saliency L(x, y) increases, the saliency difference within its neighborhood may also increase, resulting in an increase in the context modulation factor, thereby further improving the final saliency S(x, y). An increase in the global saliency G(x, y) will directly increase S(x, y). Changes in the parameters α, β, and γ will change the relative importance of local saliency, global saliency, and context modulation, affecting the final saliency map. The formula introduces a context modulation factor, which uses the significant difference between the image region and its neighborhood to enhance the contrast and highlight the truly significant area; the Gaussian kernel function is used to model the spatial distance weight, which conforms to the characteristic that the visual attention of the human eye decays with the spatial distance, and takes into account the spatial correlation and contextual information of the significance; the normalization process adapts to the dynamic range of different images, so that the algorithm can adaptively adjust to different images and improve the robustness. In summary, the formula can effectively generate clear and accurate visual saliency maps through the context modulation mechanism and multi-scale fusion strategy, providing reliable prior information for the subsequent extraction of key salient areas, which is of great significance for achieving high-quality and low-power local refresh; the parameters in the formula can be adjusted according to the actual application scenarios and system requirements to obtain the optimal saliency detection performance.

[0138] Specifically, a multi-scale and multi-feature approach is used to extract visual features of the displayed content, including color, brightness, texture, edge, motion and other features; color features can be represented by descriptors such as color histogram and color moment, reflecting the color distribution and main tone of the image. Brightness features can be represented by descriptors such as brightness histogram and brightness mean, reflecting the light and dark contrast and exposure level of the image. Texture features can be represented by descriptors such as local binary patterns and directional gradient histogram, reflecting the texture complexity and regularity of the image. Edge features can be extracted by operators such as Canny and Sobel, reflecting the contour and structural information of objects in the image. Motion features can be estimated by methods such as optical flow and block matching, reflecting the dynamic change characteristics of the image content. By extracting these visual features at multiple scales, the visual attributes of the displayed content can be fully characterized, providing rich perceptual information for subsequent saliency analysis. Based on the extracted multi-scale visual features, local saliency maps and global saliency maps are calculated. The local saliency map reflects the saliency difference between the image region and its local neighborhood, and can be calculated by methods such as the center-surround difference algorithm and the local contrast algorithm. For example, the center-surround difference algorithm takes the image patch as the center and calculates the difference between it and the surrounding patches in different visual features. The greater the difference, the higher the local significance of the center patch. The global saliency map reflects the saliency of the image area in the entire image, and can be calculated by frequency domain analysis, graph theory analysis and other methods. For example, the frequency domain analysis method extracts the global frequency domain features of the image by Fourier transforming the image, and determines the salient area according to the energy of the frequency domain coefficient. The graph theory analysis method models the image as an undirected weighted graph, and determines the global significance of the node by calculating the centrality, clustering coefficient and other indicators of the nodes in the graph. The obtained visual saliency map is threshold segmented to extract the binary mask of the salient area. Threshold segmentation is a simple and effective binarization method. By setting the saliency threshold, the saliency map is converted into a binary mask of the foreground (salient area) and the background (non-salient area). The selection of the threshold needs to balance the integrity of the salient area and the cleanliness of the background. The optimal threshold can be automatically determined by adaptive threshold selection methods such as the Otsu algorithm and the maximum entropy algorithm, or a fixed threshold can be manually set according to application requirements. After segmentation, pixels with a significance greater than a threshold are marked as key salient regions, and pixels with a significance less than or equal to the threshold are marked as non-key salient regions, thereby obtaining key salient region masks and non-key salient region masks.

[0139] Through visual saliency analysis, key areas in the displayed content can be automatically and quickly located. These areas often contain the visual elements that attract the most user attention and have a greater impact on the user's perceptual experience. Compared with resolution partitioning, visual saliency areas can more accurately and finely reflect the importance distribution of displayed content, providing a more targeted optimization basis for subsequent local refreshes. According to the distribution of key saliency areas and non-key saliency areas, while ensuring the display quality of key information, the power consumption overhead of non-key areas can be further reduced, achieving smarter and more sophisticated display optimization. At the same time, because visual saliency analysis is based on the semantic understanding of the displayed content, it can adapt to the display needs of different scenarios and different users, and has better adaptability and universality.

[0140] For example, a screenshot of a game interface is used to illustrate the process and results of visual saliency analysis:

[0141] Input image: a screenshot of a game interface, including characters, monsters, background, text and other visual elements;

[0142] Multi-scale visual feature extraction: extract visual features such as color histogram, brightness contrast, directional gradient histogram texture, Canny edge, motion vector, etc. at multiple scales such as pixel, superpixel, semantic region, etc.

[0143] Saliency calculation: Based on the extracted features, the center-surround difference algorithm is used to calculate the local saliency map, and the frequency domain analysis method is used to calculate the global saliency map;

[0144] Weighted fusion: According to the content complexity of the game interface, the local saliency weight is set to 0.7 and the global saliency weight is set to 0.3. According to the above formula for calculating the visual saliency value, the two saliency maps are weightedly superimposed to obtain the visual saliency map.

[0145] Threshold segmentation: The Otsu algorithm is used to automatically calculate the optimal significance threshold of 0.6. The areas with significance greater than 0.6 are marked as key significant areas, and the areas with significance less than or equal to 0.6 are marked as non-key significant areas, forming a binary mask.

[0146] Analysis results: In the game interface, elements such as characters, monsters, and health bars are marked as key salient areas, while elements such as backgrounds and maps are marked as non-key salient areas; this is consistent with the player's visual attention distribution, and the key salient areas cover the core interactive information of the game.

[0147] Step S3200, performing local refresh in different resolution areas, wherein the refresh frequency of the key salient area is higher than the refresh frequency of the non-key salient area;

[0148] Further, step S3200 includes:

[0149] Step S3210: In the high-resolution region, the first high-frequency F is used for the key salient region. H1 Perform local refresh and use the first low frequency F for non-critical salient areas L1 Perform a partial refresh; among them, FPS H ≤F L1 <F H1 ;

[0150] Step S3220: In the medium resolution area, the second highest frequency F is used for the key salient area. H2 Perform local refresh and use the second lowest frequency F for non-critical salient areas L2 Perform a partial refresh; among them, FPS M ≤F L2 <F H2 <FPS H ;

[0151] Step S3230: In the low-resolution region, the third low-frequency F is used for the key salient region. L3 Perform local refresh and use the fourth lowest frequency F for non-critical significant areas L4 Perform a partial refresh; where F L4 <F L3 <FPS L .

[0152] Specifically, in each resolution region, the refresh rate of the key salient region is higher than that of the non-key salient region, so as to ensure that the visually salient content has a higher display priority and temporal resolution. H1 and F L1 The value is not less than FPS H , meeting the high refresh rate requirements of high-resolution areas, while F H1 Higher than F L1 , so that the key and significant content in the high-resolution area has higher dynamic display performance and attracts the user's attention. H2 and F L2 The value is between FPS M and FPS H The display quality and power consumption are balanced, and F H2 Higher than F L2 , ensuring the dynamic display priority of key and significant content in the medium-resolution area. In the low-resolution area, F L3 and F L4 The values ​​are all lower than FPS L , minimize the power consumption in non-concern areas, but F L3 Higher than F L4, so that the dynamic display performance of key salient content in the low-resolution area is slightly better than that of non-key salient content.

[0153] By controlling the refresh rate of local areas in different resolution areas in the above-mentioned hierarchical manner, limited display resources and energy consumption budgets can be concentrated, and the display quality and dynamic performance of visually important areas can be prioritized to improve the user's visual experience. At the same time, compared with traditional global refresh, local control of refresh rate can avoid unnecessary over-refresh and reduce energy consumption under the premise of perceptual effect.

[0154] For example, suppose that in a display scene, the high-resolution area displays the main content of a video, in which the movement of the face and hands of the characters is more significant, and they belong to the key salient area; the movement of the background area is less, and it belongs to the non-key salient area. By setting the refresh rate of the face and hand areas to F H1 (such as 120Hz), set the refresh rate of the background area to F L1 (such as 90Hz), which can better display the dynamic details of facial expressions and gestures of characters, while moderately reducing the refresh rate of the background area to control power consumption. The medium-resolution area displays the subtitle information of the video. The appearance and disappearance of subtitles are key significant events, while the content of subtitles is a non-key significant area. The refresh rate of subtitle appearance and disappearance can be set to F H2 (such as 85Hz), set the refresh rate of the subtitle content to F L2 (such as 60Hz), while ensuring the smoothness of subtitle dynamic changes, reduce the power consumption of static subtitle content. The low-resolution area displays non-key information such as the video border and time code. The change of the time code belongs to the key significant area, and the static border belongs to the non-key area. The refresh rate of the time code can be set to F L3 (such as 30Hz), set the refresh rate of the border to F L4 (such as 15Hz), ensuring basic readability of the time code with very low power consumption.

[0155] In summary, step S3000 can maximize the use of the visual attention characteristics of the human eye by perceiving the saliency of the visual content and matching the refresh frequency levels of different resolution areas to perform refined refresh frequency control on local areas, and focus display resources and energy consumption on improving the display quality and dynamic performance of key areas, while non-key areas are moderately down-clocked to balance power consumption. Compared with global control, this local adaptive control strategy can provide a better visual experience under the same power consumption budget, or save more power consumption under the same visual experience level. The visual perception optimization method of the present invention has a strong practical value and is particularly suitable for mobile display devices with high requirements for energy efficiency and visual experience.

[0156] Example 2

[0157] This embodiment provides a system for reducing the power consumption of a timing control chip of a liquid crystal panel based on the embodiment 1, such as Figure 4 As shown, including:

[0158] Real-time data collection module: used to collect user's visual attention data and display content data in real time;

[0159] Visual preference prediction module: used to obtain the user's first historical viewing data, build and train the user's visual preference prediction model; determine the first focus area ROI1 and the second focus area ROI2 on the LCD panel according to the visual attention data and the display content data, combined with the user's visual preference prediction model;

[0160] Resolution area division module: used to divide the LCD panel into high-resolution area, medium-resolution area and low-resolution area;

[0161] Dynamic adjustment module: used to set corresponding pixel density thresholds and refresh rate thresholds for different resolution areas, and dynamically adjust the pixel density and refresh rate in each resolution area;

[0162] Visual saliency analysis module: used to perform visual saliency analysis on the display content of the LCD panel and determine the key saliency area and the non-key saliency area;

[0163] Local refresh module: used to perform local refresh in different resolution areas, where the refresh frequency of key salient areas is higher than that of non-key salient areas.

[0164] Example 3

[0165] This embodiment discloses a chip, including the system for reducing the power consumption of a liquid crystal panel timing control chip as described above.

[0166] Example 4

[0167] This embodiment discloses an electronic device, including the chip, memory, central processing unit and a computer program stored in the memory and executable on the central processing unit as described above. When the central processing unit executes the computer program, the method for reducing the power consumption of the liquid crystal panel timing control chip provided above is implemented.

[0168] Since the electronic device introduced in this embodiment is an electronic device used to implement the method for reducing the power consumption of the timing control chip of the liquid crystal panel in the embodiment of the present application, based on the method for reducing the power consumption of the timing control chip of the liquid crystal panel introduced in the embodiment of the present application, the technical personnel of the field can understand the specific implementation of the electronic device of the present embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not described in detail here. As long as the electronic device used by the technical personnel of the field to implement the method for reducing the power consumption of the timing control chip of the liquid crystal panel in the embodiment of the present application, it belongs to the scope of protection of this application.

[0169] Example 5

[0170] This embodiment discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the method for reducing the power consumption of a timing control chip of a liquid crystal panel is implemented.

[0171] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0172] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD) or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0173] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0174] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0175] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0176] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0177] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0178] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

[0179] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for reducing power consumption of a timing control chip of a liquid crystal panel, characterized in that: The method comprises: Acquire the user's first historical viewing data, build and train a user visual preference prediction model; collect the user's visual attention data and display content data in real time, and determine the first focus area ROI1 and the second focus area ROI2 on the liquid crystal panel in combination with the user's visual preference prediction model; Divide the liquid crystal panel into high-resolution area, medium-resolution area and low-resolution area, set corresponding pixel density thresholds and refresh frequency thresholds for different resolution areas, and dynamically adjust the pixel density and refresh frequency in each resolution area; Performing visual saliency analysis on the display content of the liquid crystal panel to determine key salient areas and non-key salient areas; performing local refresh in different resolution areas, wherein the refresh frequency of the key salient areas is higher than the refresh frequency of the non-key salient areas; The setting of corresponding pixel density thresholds and refresh frequency thresholds for different resolution areas includes: Set the pixel density lower limit of the high-resolution area to PPI H , the refresh rate is limited to FPS H ; Set the pixel density lower limit of the medium resolution area to PPI M , capped at PPI H , the refresh rate is limited to FPS M ; Set the pixel density limit of low-resolution areas to PPI L , the refresh rate is capped at FPS L ; Among them, PPI L <PPI AV <PPI M <PPI H , PPI AV is the average pixel density of the LCD panel, FPS L <FPS M <FPS H ; The dynamically adjusting the pixel density and refresh rate in each resolution area includes: Adaptively adjust the pixel density P in the high-resolution area H , satisfying P H ≥PPI H ; Adaptive adjustment of refresh rate F H , satisfying F H ≥FPS H ; Adaptively adjust the pixel density P in the medium resolution area M , meet PPI M ≤P M <PPI H ; Adaptive adjustment of refresh rate F M , meet FPS M ≤F M <FPS H ; Adaptively adjust the pixel density P in the low-resolution area L , satisfying P L <PPI L ; Adaptive adjustment of refresh rate F L , satisfying F L <FPS L ; The pixel density P is adaptively adjusted in the high-resolution area. H and refresh rate F H The methods include: Among them, A ROI1 is the area of ​​the first region of interest ROI1, A Total is the total display area of ​​the LCD panel, T ROI1 is the user's gaze time on the first region of interest ROI1, T Total is the total gaze time of the user on the entire LCD panel, α1 is the pixel density sensitivity coefficient, and β1 is the refresh rate sensitivity coefficient.

2. The method for reducing power consumption of a timing control chip of a liquid crystal panel according to claim 1, characterized in that: The determining of the first region of interest ROI1 and the second region of interest ROI2 on the liquid crystal panel comprises: Input the real-time collected visual attention data and display content data into the user visual preference prediction model to obtain the predicted visual preference distribution map and attention area coordinates; According to the predicted visual preference distribution map and the coordinates of the regions of interest, a first region of interest ROI1 and a second region of interest ROI2 on the liquid crystal panel are determined.

3. The method for reducing power consumption of a timing control chip of a liquid crystal panel according to claim 1, characterized in that: The method for dividing the liquid crystal panel into high-resolution area, medium-resolution area and low-resolution area is: dividing the area where ROI1 is located into a high-resolution area, dividing the area where ROI2 is located into a medium-resolution area, and dividing the non-ROI area into a low-resolution area.

4. The method for reducing power consumption of a liquid crystal panel timing control chip according to claim 3, characterized in that: The region where ROI1 is located is divided into a high-resolution region by taking the center of ROI1 as the center of the circle and the size of ROI1 as the radius to divide a circular or elliptical high-resolution region; the high-resolution region completely covers ROI1; The region where ROI2 is located is divided into a medium-resolution region by taking the center of ROI2 as the center of the circle and the size of ROI2 as the radius to divide a circular or elliptical medium-resolution region; the medium-resolution region completely covers ROI2.

5. The method for reducing power consumption of a liquid crystal panel timing control chip according to claim 1, characterized in that: Determining the key salient areas and non-key salient areas includes: Extract multi-scale visual features of the content displayed on the LCD panel; Based on the extracted multi-scale visual features, the local saliency map and the global saliency map of the displayed content are calculated; Generate the final visual saliency map based on the local saliency map and the global saliency map; A significance threshold is set. The areas in the visual significance map whose significance is greater than the significance threshold are marked as key significant areas, and the areas whose significance is less than or equal to the significance threshold are marked as non-key significant areas.

6. The method for reducing power consumption of a timing control chip of a liquid crystal panel according to claim 1, characterized in that: The local refreshing in different resolution areas includes: In the high-resolution area, the first high frequency F is used for the key salient area. H1 Perform local refresh and use the first low frequency F for non-critical salient areas L1 Perform a partial refresh; among them, FPS H ≤F L1 <F H1 ; In the medium resolution area, the second highest frequency F is used for the key salient area. H2 Perform local refresh and use the second lowest frequency F for non-critical salient areas L2 Perform a partial refresh; among them, FPS M ≤F L2 <F H2 <FPS H ; In the low-resolution area, the third lowest frequency F is used for the key salient area. L3 Perform local refresh and use the fourth lowest frequency F for non-critical significant areas L4 Perform a partial refresh; where F L4 <F L3 <FPS L .

7. A system for reducing power consumption of a timing control chip for a liquid crystal panel, which is used to implement the method for reducing power consumption of a timing control chip for a liquid crystal panel according to any one of claims 1 to 6, characterized in that: The system comprises: Real-time data collection module: used to collect user's visual attention data and display content data in real time; Visual preference prediction module: used to obtain the user's first historical viewing data, build and train the user's visual preference prediction model; determine the first focus area ROI1 and the second focus area ROI2 on the LCD panel according to the visual attention data and the display content data, combined with the user's visual preference prediction model; Resolution area division module: used to divide the LCD panel into high-resolution area, medium-resolution area and low-resolution area; Dynamic adjustment module: used to set corresponding pixel density thresholds and refresh rate thresholds for different resolution areas, and dynamically adjust the pixel density and refresh rate in each resolution area; Visual saliency analysis module: used to perform visual saliency analysis on the display content of the LCD panel and determine the key saliency area and the non-key saliency area; Local refresh module: used to perform local refresh in different resolution areas, where the refresh frequency of key salient areas is higher than that of non-key salient areas.

8. A chip, characterized in that: It comprises the system for reducing the power consumption of a timing control chip of a liquid crystal panel as described in claim 7.

Images