A medical image display system based on color calibration

Abstract

The present application relates to the technical field of image processing, and particularly relates to a medical image display system based on color calibration, comprising: a calibration tensor generation module: generating a three-dimensional calibration input tensor; a closed-loop feedback calibration module: inputting the three-dimensional calibration input tensor into a pre-trained multi-modal adversarial compensation network to generate a pixel-level aging-environment coupling compensation coefficient matrix and a color gamut anchor point coordinate; and a compensation execution module: writing a compensated quantized color instruction set into a redundant driving unit of a display panel to realize collaborative cancellation of device aging and environmental interference. The present application can effectively reduce color restoration errors in high-resolution medical images, improve color stability, and is suitable for high-precision diagnostic scenarios such as real reproduction of endoscopic and CT pseudo-color images.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to a medical image display system based on color calibration. BACKGROUND

[0002] As an important information carrier for clinical diagnosis and intraoperative navigation, medical images have very high requirements for the color restoration accuracy and stability of their display systems. Currently, medical display devices generally use high-resolution panels such as liquid crystal (LCD), organic light-emitting diode (OLED), or quantum dot light-emitting diode (QLED). However, in the long-term operation process, there is an irreversible brightness decay problem in each sub-pixel of the panel, and the aging rates of different materials are inconsistent in the red, green, and blue channels, leading to increasingly serious color imbalance.

[0003] On the other hand, the environmental light sources in medical scenarios such as operating rooms and radiology reading rooms are complex and variable, and their reflected and scattered light can significantly interfere with the display content. Especially when displaying low-contrast images (such as angiography and pathological sections), problems such as false color, color deviation, and reduced contrast often occur, which seriously affect the accuracy of diagnosis and visual comfort. Although some existing display systems integrate brightness sensors or ambient light sensors, most of them only perform rough adjustment on the overall brightness and cannot achieve fine compensation for specific display areas and specific pixels.

[0004] In addition, traditional color calibration techniques are mostly based on preset color lookup tables (LUTs) or offline correction algorithms, which are difficult to dynamically adapt to the coordinated changes of device aging and environmental interference. Moreover, the compensation parameters are usually generated by single-channel models, which lack deep perception of image content characteristics, leading to defects such as color gamut drift, inter-frame inconsistency, and response lag in the compensated images, especially in dynamic medical image sequences. SUMMARY

[0005] The present application provides a medical image display system based on color calibration, which realizes sub-pixel level dynamic color correction through a multi-modal color compensation mechanism of device aging state, environmental light interference characteristics, and image content color gamut distribution, combined with intelligent feedback and hardware driving collaboration, to ensure the color accuracy and system stability of medical images in complex application scenarios.

[0006] A medical image display system based on color calibration, comprising:

[0007] A calibration tensor generation module: synchronously acquires real-time aging fingerprint spectrum, environmental light interference matrix, and color gamut feature topology of the image to be displayed, and generates a three-dimensional calibration input tensor, wherein the aging fingerprint spectrum includes the brightness decay gradient distribution of each sub-pixel of the display panel.

[0008] Closed-loop feedback calibration module: The three-dimensional calibration input tensor is input into a pre-trained multimodal adversarial compensation network to generate a pixel-level aging-environment coupling compensation coefficient matrix and color gamut anchor coordinates. The compensation coefficient matrix and color gamut anchor coordinates form a closed-loop feedback calibration path.

[0009] Compensation execution module: Based on the closed-loop feedback calibration path, the input image is remapped at the subpixel level, and the compensated quantum color instruction set is written into the redundant drive unit of the display panel to achieve the coordinated cancellation of equipment aging and environmental interference.

[0010] Optionally, the calibration tensor generation module specifically includes:

[0011] Aging fingerprint pattern construction unit: The luminous efficiency decay value of red, green and blue sub-pixels is detected pixel by pixel in a frame scanning manner through a micro photoelectric sensor array embedded in the display panel, and the aging fingerprint pattern is constructed based on the spatial distribution of the decay value;

[0012] Ambient light interference matrix generation unit: A multispectral ambient light probe is used to synchronously collect the wavelength-intensity interferogram of ambient light. The coherence characteristics of light intensity in each band are extracted by Fourier transform to generate an ambient light interference matrix including spectral offset and intensity modulation factor.

[0013] Color Gamut Feature Topology Analysis Unit: Performs color gamut feature analysis on the image to be displayed, extracts the topological relationship between the chromaticity coordinates and luminance values ​​of each color block in the image in the CIE xyY space, and constructs a color gamut feature topology map;

[0014] Three-dimensional calibration tensor construction unit: The spatial attenuation distribution of the aging fingerprint spectrum, the frequency domain feature vector of the ambient light interference matrix, and the chromaticity correlation weight of the color gamut feature topology map are spliced ​​into a three-dimensional calibration input tensor.

[0015] Optionally, in the aging fingerprint spectrum, the brightness attenuation gradient distribution of each sub-pixel is represented as:

[0016] ;

[0017] in, Indicates the first Location of sub-pixel type The normalized brightness attenuation gradient, Indicates subpixel type The initial brightness value when it left the factory. Indicates the current time. The brightness value of the sub-pixel. Indicates cumulative working time The relevant nonlinear aging coefficients, formally represented by the aging fingerprint spectrum, are as follows: .

[0018] Optionally, the three-dimensional calibration input tensor is formed along the channel dimension, with a size of H×W×(N+M+P), where H and W are the display panel resolution, N is the number of aging parameter channels, M is the ambient light feature dimension, and P is the color gamut topology correlation dimension.

[0019] Optionally, the multimodal adversarial compensation network includes a generator branch and a compensation verifier branch, and the closed-loop feedback calibration module specifically includes:

[0020] Compensation coefficient generation unit: The generator branch uses depthwise separable convolution to extract the aging-environment coupling features in the three-dimensional calibration input tensor, and generates a pixel-level aging-environment coupling compensation coefficient matrix including red, green and blue sub-pixel compensation gain values;

[0021] Color Gamut Anchor Point Extraction Unit: The compensation verifier branch synchronously processes the color gamut feature topology map in the three-dimensional calibration input tensor, extracts the dynamic distribution features of the image color gamut through an adaptive spatial pooling layer, and outputs a set of color gamut anchor point coordinates in the CIE LAB color space;

[0022] Generator-Verifier Adversarial Unit: The generator branch and the compensation verifier branch form a dynamic game mechanism through adversarial training: The compensation coefficient matrix is ​​applied to the test image under the simulated aging environment, and the compensated image is input into the compensation verifier branch for color gamut anchor point matching degree verification. If the color gamut anchor point coordinate offset exceeds the preset threshold, the convolution kernel weight of the generator branch is corrected in reverse.

[0023] Closed-loop calibration path construction unit: Performs spatiotemporal correlation analysis on the coordinates of the color gamut anchor point generated in the current frame and the color gamut feature topology of the next frame image, generates the color gamut anchor point drift vector and feeds it back to the generator branch, and adjusts the spatial distribution weight of the compensation coefficient matrix in real time.

[0024] Nonlinear mapping unit for compensation coefficients: The pixel-level aging-environment coupling compensation coefficient matrix forms a nonlinear mapping relationship with the color gamut anchor point coordinates.

[0025] Optionally, the adaptive spatial pooling layer extracts the dynamic distribution features of the image color gamut and outputs a set of color gamut anchor point coordinates in the CIE LAB color space, specifically including:

[0026] After receiving the gamut feature topology map from the calibration input tensor, the compensation verifier branch enhances the key color response regions in the gamut channels through the channel attention mechanism to obtain the enhanced gamut feature map. The enhanced gamut feature map is then input into the adaptive spatial pooling layer. The adaptive spatial pooling layer automatically adjusts the pooling window size and sliding step based on the color distribution density of different regions in the current image, enabling the pooling operation to adaptively perform fine-grained aggregation on high color change regions and large-scale compression on low change regions, thus preserving the structural information of color transition regions and edges.

[0027] The compact color gamut representation obtained after pooling is input into the feature compressor, converted to the CIE LAB color space, and clustered and extracted for each local region. The representative color center is extracted as the color gamut anchor point, which reflects the position of the main color distribution of the current frame image in the perceptual space.

[0028] Optionally, the spatiotemporal correlation analysis in the closed-loop calibration path construction unit specifically includes:

[0029] When performing color gamut anchor point drift analysis, the set of color gamut anchor point coordinates extracted from the current frame image is spatially aligned with the color gamut feature topology map in the next frame image to ensure that the correspondence of color gamut anchor points between the two frames is consistent in the display area. Feature matching is performed using the color distribution features between anchor points to determine the optimal corresponding target of the current frame anchor point in the next frame.

[0030] For each successfully matched color gamut anchor point, the direction and magnitude of color drift are determined based on the coordinate difference in the CIE LAB space and the position change in the image space. An anchor point drift vector across frames is generated and fed back to the compensation network to adjust the compensation parameter distribution of subsequent frames, thereby maintaining color consistency between consecutive frames.

[0031] Optionally, in the nonlinear mapping unit of the compensation coefficient matrix, the nonlinear mapping relationship between the pixel-level aging-environment coupling compensation coefficient matrix and the aging fingerprint spectrum attenuation gradient and the offset distance between the current color gamut anchor point and the standard color gamut is expressed in full sigmoid form as follows:

[0032] ;in, Indicates the first The attenuation gradient of the aging fingerprint pattern of the position sub-pixel. Indicates the corresponding color channel The Euclidean distance between the color gamut anchor point and the center of the standard color gamut. This represents the mapping sensitivity factor that is dynamically optimized during adversarial training.

[0033] Optionally, the compensation execution module specifically includes:

[0034] Brightness correction calculation unit: Performs sub-pixel convolution operation on the pixel-level aging-environment coupling compensation coefficient matrix and the original pixel values ​​of the current input image to generate compensated RGB three-channel brightness correction values;

[0035] Boundary constraint and coordinate remapping unit: Based on the set of color gamut anchor point coordinates, a dynamic color gamut enclosure is constructed in the CIE LAB color space, color boundary constraints are applied to the corrected pixel values, and sub-pixel level color coordinate remapping is achieved through the Lagrange interpolation algorithm;

[0036] Color instruction writing unit: converts the remapped color data into a color instruction set, which includes the driving pulse width modulation parameters and voltage gradient compensation values ​​for each sub-pixel, and writes them in real time through the spare TFT circuit in the redundant driving unit built into the display panel.

[0037] Optionally, the compensation execution module is further provided with a double buffer register, specifically including a double buffer register group in the redundant drive unit;

[0038] When the first register group executes the current frame compensation instruction, the second register group synchronously receives the closed-loop feedback calibration parameters of the next frame, realizing low-latency switching between aging compensation and environmental interference cancellation.

[0039] The beneficial effects of this invention are:

[0040] This invention constructs a three-dimensional calibration input tensor that includes display device aging fingerprint spectrum, ambient light interference matrix, and image color gamut feature topology. This tensor comprehensively perceives key factors affecting color accuracy and inputs them into a multimodal adversarial compensation network to achieve collaborative modeling and pixel-level compensation for display aging and environmental changes. Compared with traditional methods that only perform single-channel correction based on brightness or color temperature parameters, this invention can effectively reduce color reproduction errors in high-resolution medical images and improve color stability. It is suitable for high-precision diagnostic scenarios such as the realistic reproduction of endoscopic and CT pseudocolor images.

[0041] In this invention, the generator branch can dynamically output aging-environment coupling compensation coefficients, while the verifier branch extracts color gamut anchor points through adaptive spatial pooling and verifies the compensation effect in real time. The two form a closed-loop optimization path through adversarial training mechanism and anchor point drift vector feedback. This not only enables self-supervised tuning of color remapping for each frame, but also maintains color gamut spatial continuity between multiple frames, overcoming the color flickering and offset problems caused by device state fluctuations or inter-frame differences in the prior art, and improving the matching degree.

[0042] This invention introduces redundant drive units and dual-buffer register groups into the compensation execution module, and combines color instruction quantization and reverse compensation current injection mechanisms to overcome the limitations of traditional methods that rely solely on external algorithm correction. It achieves hardware-level solutions for TFT aging, sub-pixel drive attenuation, and line impedance drift. Especially in high-level medical display devices, this structure can achieve pulse width + voltage dual-channel control and quickly switch to the backup channel when the main drive fails, ensuring image continuity and color accuracy. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the system functional modules according to an embodiment of the present invention;

[0045] Figure 2 This is a schematic diagram of the calibration tensor generation module according to an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of the closed-loop feedback calibration module according to an embodiment of the present invention;

[0047] Figure 4 This is a schematic diagram of the compensation execution module in an embodiment of the present invention. Detailed Implementation

[0048] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. For some well-known technologies, those skilled in the art may also use other alternative methods to implement the invention. Moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0049] like Figures 1-4 As shown, a color-calibrated medical image display system includes:

[0050] Calibration Tensor Generation Module: Synchronously acquires the real-time aging fingerprint spectrum of the display device, the ambient light interference matrix, and the color gamut feature topology of the image to be displayed, and generates a three-dimensional calibration input tensor. The aging fingerprint spectrum includes the brightness attenuation gradient distribution of each sub-pixel of the display panel.

[0051] Closed-loop feedback calibration module: The three-dimensional calibration input tensor is input into the pre-trained multimodal adversarial compensation network to generate a pixel-level aging-environment coupling compensation coefficient matrix and color gamut anchor coordinates. The compensation coefficient matrix and color gamut anchor coordinates form a closed-loop feedback calibration path.

[0052] Compensation execution module: Based on the closed-loop feedback calibration path, the input image is remapped at the subpixel level. The compensated quantum color instruction set is written into the redundant drive unit of the display panel to achieve the coordinated cancellation of equipment aging and environmental interference.

[0053] The calibration tensor generation module specifically includes:

[0054] Aging fingerprint map construction unit: Using a miniature photoelectric sensor array embedded in the display panel, the luminous efficiency decay values ​​of red, green, and blue sub-pixels are detected pixel-by-pixel in a frame-scanning manner. An aging fingerprint map is constructed based on the spatial distribution of these decay values. The brightness decay gradient distribution of each sub-pixel is calculated using the following formula:

[0055] ;in, Indicates the first Location of sub-pixel type The normalized brightness attenuation gradient, Indicates subpixel type The initial brightness value when it left the factory. Indicates the current time. The brightness value of the sub-pixel. Indicates cumulative working time The relevant nonlinear aging coefficient is expressed as follows: ,in, Let be the aging constant of the display panel material. Assume the display panel resolution is... Then the aging fingerprint spectrum can be formally represented as: .

[0056] Ambient light interference matrix generation unit: The multispectral ambient light probe is used to synchronously collect the wavelength-intensity interferogram of ambient light, and the coherence characteristics of light intensity in each band are extracted by Fourier transform to generate an ambient light interference matrix containing spectral offset and intensity modulation factor.

[0057] Suppose that the wavelength-intensity sequence of ambient light collected by the multispectral ambient light probe is:

[0058] ; For the first The center wavelength of each band For true spectral intensity, To measure the noise interference term, The sampling sequence along the wavelength axis is calculated to represent the total number of sampling bands (e.g., dividing the 400-700nm visible light region into 32-64 segments). Perform a one-dimensional Fast Fourier Transform (FFT):

[0059] ;in, For frequency index (corresponding to the periodic components of the band), For the first The complex amplitudes of each frequency component; the coherence factor and spectral shift are defined as follows: ;in, For the first Normalized intensity factor of frequency domain components, The maximum response wavelength (i.e., the offset center) corresponding to the principal component. The real part of the complex spectrum is represented; the final ambient light interference matrix can be expressed as:

[0060] ;in It is the number of selected dominant frequency components, and each row of the matrix corresponds to a band interferometric feature (intensity factor + offset center).

[0061] Color Gamut Feature Topology Analysis Unit: Performs color gamut feature analysis on the image to be displayed, extracts the topological relationships of the chromaticity coordinates and luminance values ​​of each color block in the image in the CIE xyY space, and constructs a color gamut feature topology map. Specifically, this includes converting pixel RGB to CIExyY coordinates.

[0062] ;

[0063] For any pixel in the image to be displayed The RGB values ​​are converted to XYZ tristimulus values, where, RGB to XYZ color space conversion matrix (based on sRGB or D65 standard), This represents the CIE1931 tristimulus value. The input values ​​are normalized RGB values. Next, the x, y, and Y coordinates are calculated:

[0064] ;

[0065] The entire image is sorted by pixels. Triple clustering yields Anchor points for individual color blocks (clusters); construct a topology graph between color blocks. ,in, Each vertex represents a color patch. Edge weights are defined as the color difference between color blocks plus spatial adjacency weights.

[0066] ;in, Indicates the first The chromaticity-luminance vector at the center of each color block The spatial distance between the centers of the color block anchor points. Using Gaussian adjustment factors for color difference and spatial distance, the final constructed color gamut feature topology map is used in channel stitching. The input tensor is expanded in the form of a spectral feature map.

[0067] 3D calibration tensor building blocks: The following three types of features are fused using tensors and stitched together along the channel dimension to construct a unit with a size of [size missing]. 3D calibration input tensor:

[0068] Spatial decay distribution characteristics of aging fingerprint spectrum (channel dimension is 10 ... );

[0069] The frequency domain eigenvectors of the ambient light interference matrix (channel dimension number is...) );

[0070] Chromaticity correlation weights of gamut feature topology (channel dimension is ) );

[0071] in, , These are the vertical and horizontal resolutions of the display panel, respectively. Here, represents the number of aging parameter dimensions, and represents the number of decay layers for each sub-pixel type. The ambient light frequency domain characteristic dimension is determined based on the probe's band resolution. The correlation dimension of the color gamut topology is determined based on the image segmentation and color clustering results.

[0072] The closed-loop feedback calibration module specifically includes:

[0073] The compensation coefficient generation unit of the generator branch: The multimodal adversarial compensation network includes a generator branch and a compensation verifier branch. The generator branch receives a 3D calibration input tensor and uses a depthwise separable convolutional structure to extract the aging channel-environment channel coupling features from the tensor, generating a coefficient of size [missing information]. Pixel-level compensation coefficient matrix: ;in, Indicates subpixel type In position The compensation gain value on, This refers to the resolution dimension of the display panel.

[0074] The validator branch's color gamut anchor point extraction unit: The compensated validator branch synchronously processes the color gamut feature topology map in the calibration input tensor, extracts the dynamic color gamut distribution features of different image regions through an adaptive spatial pooling layer, and outputs a set of color gamut anchor point coordinates in the CIELAB color space.

[0075] ;in, The number of color gamut anchor points extracted from the image. Indicates the first CIELAB coordinates of each color gamut anchor point.

[0076] The adaptive spatial pooling layer is as follows:

[0077] After receiving the gamut feature topology map from the calibration input tensor, the compensation verifier branch first enhances the key color response regions in the gamut channels through a channel attention mechanism. Then, the enhanced gamut feature map is input to an adaptive spatial pooling layer. This pooling layer automatically adjusts the pooling window size and sliding step based on the color distribution density of different regions in the current image. This allows the pooling operation to adaptively perform fine-grained aggregation on high-color-variety regions and large-scale compression on low-variety regions, thus preserving the structural information of color transition regions and edges. The compact gamut representation obtained after pooling is fed into a feature compressor, converted to the CIE LAB color space, and subjected to cluster analysis and color center extraction on each local region. Finally, representative color centers are extracted as gamut anchor points. These anchor point sets reflect the main color distribution positions of the current frame image in perceptual space, serving as a reference for subsequent compensation accuracy verification.

[0078] Generator-Verifier Adversarial Unit: Closed-loop training is performed through a dynamic game mechanism, specifically:

[0079] 1. The compensation coefficient matrix The test images are applied to simulated aging environments to generate compensated images. ;

[0080] 2. Input validator branch, extract its color gamut anchor point set ;

[0081] 3. If the offset of any color gamut anchor point is:

[0082] This triggers backpropagation of the error, updating the convolutional kernel weights of the generator branch, where... The color gamut anchor point offset tolerance threshold is 1.5 CIE units. The distance is Euclidean.

[0083] Closed-loop calibration path construction unit: generates the set of anchor point coordinates in the current frame image. Color gamut topology map with the next frame image Perform spatiotemporal correlation analysis and calculate the anchor point drift vector. :

[0084] The anchor point drift vector is fed back to the generator branch, and the spatial distribution weights of the compensation coefficient matrix are adjusted in real time during the next training round.

[0085] The spatiotemporal correlation analysis specifically involves the following steps: During anchor point drift analysis, the set of color gamut anchor point coordinates extracted from the current frame image is first spatially aligned with the color gamut feature topology map in the next frame image. This ensures consistent correspondence between anchor points in the display area between the two frames. Subsequently, feature matching is performed using color distribution features between anchor points (such as chromaticity values, brightness levels, and local adjacency) to determine the optimal corresponding target for the current frame anchor point in the next frame. For each successfully matched anchor point pair, the direction and magnitude of the color drift are comprehensively judged based on their coordinate differences in CIE LAB space and their positional changes in image space. This generates a cross-frame anchor point drift vector, reflecting both the display offset trend caused by device aging or environmental changes and the impact of dynamic changes in image content on color gamut distribution. Finally, this drift vector is fed back to the compensation network to dynamically adjust the compensation parameter distribution of subsequent frames, achieving color consistency maintenance between consecutive frames.

[0086] Pixel-level compensation coefficient nonlinear mapping unit: pixel-level aging-environment coupling compensation coefficient matrix Attenuation gradient of aging fingerprint spectrum and the offset distance between the current color gamut anchor point and the standard color gamut The following nonlinear mapping relationship of sigmoid form exists between them:

[0087] ;in, Indicates the first The decay gradient of the position sub-pixel Indicates the corresponding color channel The Euclidean distance between the color gamut anchor point and the center of the standard color gamut. This represents the mapping sensitivity factor that is dynamically optimized during adversarial training.

[0088] The compensation execution module specifically includes:

[0089] Brightness correction calculation unit: calculates the pixel-level aging-environment coupling compensation coefficient matrix. Compared with the original pixel values ​​of the current input image Perform sub-pixel convolution processing to obtain the brightness correction value for each sub-pixel channel. Taking the red channel as an example, the correction calculation is as follows:

[0090] ;

[0091] in, The corrected brightness value for the red sub-pixel. Indicates the original red channel brightness. This represents the aging-environment coupling compensation gain coefficient. This represents the decay gradient of the red sub-pixels in the aging fingerprint pattern. This represents the Euclidean distance between the current red channel's color gamut anchor point and the center of the standard color gamut. It is a very small constant used to prevent division by zero; other channels... Same processing as above.

[0092] Boundary Constraint and Coordinate Remapping Unit: Based on the extracted set of color gamut anchor point coordinates, a minimum bounding box is constructed in the CIE LAB color space to limit the color gamut boundary range of the calibrated pixel values. For the corrected pixel values ​​that fall within the boundary range, the Lagrange interpolation algorithm is used to interpolate and remap their chromaticity coordinates between anchor points to achieve sub-pixel-level accurate color positioning and avoid excessive offset or color jump.

[0093] Color instruction writing unit: quantizes the remapped pixel data into a color driving instruction set, which includes:

[0094] Pulse width modulation (PWM) parameters for each sub-pixel;

[0095] The compensation value corresponding to the voltage gradient;

[0096] The above instructions are written in real time to the backup TFT circuit of the redundant drive unit in the display panel, and are updated frame by frame by the hardware controller.

[0097] Double-buffered registers (for low-latency switching):

[0098] A dual-buffered register set is configured in the redundant drive unit:

[0099] The first register group is used for real-time driving of the color compensation parameters for the current frame;

[0100] The second register group receives the calibration parameters output by the compensation network closed loop in the background for the next frame;

[0101] The two register groups rotate alternately to achieve seamless switching between compensation instruction frames, ensuring no flickering or stuttering in high frame rate display environments.

[0102] It also includes redundant routing and reverse compensation current injection mechanisms: When the main drive circuit detects faults such as abnormal resistance or delay offset, the control module triggers routing switching logic to switch the sub-pixel electrode connection to redundant routing, and at the same time starts reverse compensation current injection, the current intensity of which is defined as follows:

[0103] ,in, This represents the compensation current injected in reverse. The electroluminescent efficiency coefficient of the display panel material (depending on the physical characteristics of OLED or QLED devices), The rate of change of the current sub-pixel decay gradient over time. This represents the Euclidean distance between the corresponding channel color gamut anchor point and the standard color center. This mechanism is used to quickly neutralize trace charge accumulation, improve pixel response speed, and delay electrode material aging.

[0104] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0105] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A color calibration based medical image display system, characterized by, The application relates to a display device calibration method and device. The calibration tensor generation module synchronously obtains a real-time aging fingerprint spectrum of a display device, an ambient light interference matrix and a color gamut feature topology of a to-be-displayed image, generates a three-dimensional calibration input tensor, and the aging fingerprint spectrum comprises a luminance attenuation gradient distribution of each sub-pixel of a display panel; The closed-loop feedback calibration module inputs the three-dimensional calibration input tensor into a pre-trained multi-modal adversarial compensation network, generates a pixel-level aging-environment coupling compensation coefficient matrix and a color gamut anchor point coordinate, and the compensation coefficient matrix and the color gamut anchor point coordinate form a closed-loop feedback calibration path; The compensation execution module performs sub-pixel level color remapping on the input image based on the closed-loop feedback calibration path, writes a compensated quantized color instruction set into a redundant driving unit of the display panel, and realizes collaborative compensation of device aging and environmental interference; The multi-modal adversarial compensation network comprises a generator branch and a compensation verifier branch, and the closed-loop feedback calibration module specifically comprises: The compensation coefficient generation unit: the generator branch adopts a depth separable convolution to extract aging-environment coupling features in the three-dimensional calibration input tensor, and generates a pixel-level aging-environment coupling compensation coefficient matrix comprising red, green and blue sub-pixel compensation gain values; The color gamut anchor point extraction unit: the compensation verifier branch synchronously processes the color gamut feature topology in the three-dimensional calibration input tensor, extracts dynamic distribution features of image color gamut through an adaptive spatial pooling layer, and outputs a color gamut anchor point coordinate set in a CIE LAB color space; The generator-verifier adversarial unit: the generator branch and the compensation verifier branch form a dynamic game mechanism through adversarial training: the compensation coefficient matrix is applied to a test image in a simulated aging environment to generate a compensated image, the compensated image is input into the compensation verifier branch for color gamut anchor point matching degree verification, and if a color gamut anchor point coordinate offset exceeds a preset threshold, the convolution kernel weight of the generator branch is reversely corrected; The closed-loop calibration path construction unit: the color gamut anchor point coordinates generated in a current frame are subjected to space-time correlation analysis with the color gamut feature topology of a next frame image, a color gamut anchor point drift vector is generated and fed back to the generator branch, and the spatial distribution weight of the compensation coefficient matrix is adjusted in real time; The compensation coefficient nonlinear mapping unit: the pixel-level aging-environment coupling compensation coefficient matrix and the color gamut anchor point coordinate form a nonlinear mapping relationship.

2. The color calibration based medical image display system of claim 1, wherein, The calibration tensor generation module specifically comprises: The aging fingerprint spectrum construction unit: a micro photoelectric sensor array embedded in the display panel is used to detect the light-emitting efficiency decay value of red, green and blue sub-pixels in a frame scanning mode, and the aging fingerprint spectrum is constructed based on the spatial distribution of the decay value; The ambient light interference matrix generation unit: a multi-spectral ambient light probe is used to synchronously collect a wavelength-intensity interference spectrum of ambient light, the coherence features of light intensity of various wavebands are extracted through Fourier transform, and an ambient light interference matrix comprising a spectral shift and an intensity modulation factor is generated; The color gamut feature topology analysis unit: color gamut feature analysis is performed on the to-be-displayed image, the topological correlation relationship of the chrominance coordinates and the luminance values of each color block in the image in the CIExyY space is extracted, and a color gamut feature topology graph is constructed. The three-dimensional calibration tensor construction unit: tensor splicing is performed on the spatial decay distribution of the aging fingerprint, the frequency domain feature vector of the ambient light interference matrix, and the chrominance correlation weight of the color gamut feature topology, to generate a three-dimensional calibration input tensor.

3. The color calibration based medical image display system of claim 2, wherein, In the aging fingerprint, the luminance decay gradient distribution of each sub-pixel is represented as: ； wherein, denotes the normalized luminance decay gradient of the sub-pixel type at the position, denotes the luminance value of the sub-pixel type at initial factory, denotes the luminance value of the sub-pixel at the current time, denotes the non-linear aging coefficient related to the cumulative working time and the aging fingerprint is formalized as: wherein, H, W are the resolution of the display panel.

4. The color calibration based medical image display system of claim 2, wherein, The three-dimensional calibration input tensor is formed along the channel dimension and has a size of HxWx(N+M+P), where H and W are the display panel resolution, N is the number of aging parameter channels, M is the ambient light feature dimension, and P is the color gamut topology correlation dimension.

5. The color calibration based medical image display system of claim 1, wherein, The dynamic distribution feature of the image color gamut output by the adaptive spatial pooling layer includes a set of color gamut anchor point coordinates in the CIE LAB color space, which specifically includes: After the compensation verifier branch receives the color gamut feature topology in the calibration input tensor, the channel attention mechanism is used to enhance the key color response area in the color gamut channel, to obtain an enhanced color gamut feature map. The enhanced color gamut feature map is input into the adaptive spatial pooling layer, which automatically adjusts the pooling window size and sliding step based on the color distribution density of different regions in the current image, so that the pooling operation adaptively performs fine-grained aggregation on high color change regions and performs large-scale compression on low change regions, while retaining the structural information of the color transition region and the edge. The compact color gamut representation obtained after the pooling is input into the feature compressor, which is converted to the CIE LAB color space and clustered and color center extracted for each local region. The representative color centers are extracted as color gamut anchor points, which reflect the main color distribution position of the current frame image in the perception space.

6. The color calibration based medical image display system of claim 1, wherein, The spatio-temporal correlation analysis in the closed-loop calibration path construction unit specifically includes: When performing color gamut anchor point drift analysis, the set of color gamut anchor point coordinates extracted from the current frame image is spatially aligned with the color gamut feature topology in the next frame image, to ensure that the corresponding relationship of the color gamut anchor points between the two frames is consistent. Feature matching is performed using the color distribution feature between the anchor points to determine the optimal corresponding target of the current frame anchor point in the next frame. For each pair of successfully matched color gamut anchor points, the direction and amplitude of the color drift are determined based on the coordinate difference in the CIE LAB space and the position change in the image space, to generate an anchor point drift vector across frames. The anchor point drift vector is fed back to the compensation network to adjust the compensation parameter distribution of the subsequent frame, thereby maintaining color consistency between consecutive frames.

7. The color calibration based medical image display system of claim 1, wherein, In the compensation coefficient nonlinear mapping unit, the pixel-level aging-environment coupling compensation coefficient matrix satisfies a nonlinear mapping relationship in the sigmoid form with the aging fingerprint decay gradient and the offset distance between the current color gamut anchor point and the standard color gamut, which is represented as: ; wherein, represents the normalized luminance decay gradient of the subpixel type at the Euclidean distance of the color gamut anchor point corresponding to the subpixel type from the standard color gamut center, represents the mapping sensitivity factor dynamically optimized in the adversarial training.

8. The color calibration based medical image display system of claim 1, wherein, The compensation execution module specifically includes: The luminance correction calculation unit: performs sub-pixel convolution operation on the pixel-level aging-environment coupling compensation coefficient matrix and the original pixel value of the current input image, to generate compensated RGB three-channel luminance correction values. The boundary constraint and coordinate remapping unit: according to the color gamut anchor point coordinate set, a dynamic color gamut enclosure is constructed in the CIE LAB color space, the corrected pixel value is subjected to chromaticity boundary constraint, and sub-pixel level color coordinate remapping is realized through a Lagrange interpolation algorithm; The color instruction writing unit: the remapped color data is converted into a color instruction set, the color instruction set includes driving pulse width modulation parameters and voltage gradient compensation values of each sub-pixel, and is written in real time through a spare TFT circuit in a redundant driving unit built in the display panel.

9. The color calibration based medical image display system of claim 1, wherein, The compensation execution module is further provided with a double buffering register, specifically including a double buffering register group arranged in the redundant driving unit; When the first register group executes the current frame compensation instruction, the second register group synchronously receives the closed loop feedback calibration parameters of the next frame, so that low delay switching of aging compensation and environmental interference cancellation is realized.

Images