A method for detecting heat dissipation performance of a heat sink on a heat dissipation object

Abstract

The application provides a detection method for heat dissipation performance of a heat dissipation object by a heat dissipation device, and belongs to the technical field of heat dissipation performance detection. The detection method comprises the following steps: smearing a coloring agent on a first preset surface, then adhering the heat dissipation device and the heat dissipation object according to a heat dissipation installation position; heating the adhered heat dissipation device and heat dissipation object to dry the coloring agent; separating the heat dissipation device and the heat dissipation object, then obtaining a detection index, and finally judging the heat dissipation performance of the heat dissipation device to the heat dissipation object according to the detection index. One of the contact surface of the heat dissipation device and the heat dissipation object and the heat dissipation object is the first preset surface, and the other is the second preset surface; the detection index is used to represent the decolorization of the first preset surface and / or the coloring of the second preset surface. The detection method can accurately reflect the contact between the heat dissipation device and the heat dissipation object through the detection index, so as to effectively reflect the actual heat dissipation effect of the heat dissipation device to the heat dissipation object.

Description

Technical Field

[0001] This application relates to the field of heat dissipation performance testing technology, and more specifically, to a method for testing the heat dissipation performance of a radiator on a heat dissipation object. Background Technology

[0002] Currently, methods for testing the thermal conductivity and heat transfer performance of heat sinks on electronic devices typically only assess the heat dissipation characteristics of the heat sink itself. These methods cannot effectively reflect the actual heat dissipation effect that the heat sink can exert on chips and other electronic devices, which is not conducive to accurate thermal control of electronic devices. Summary of the Invention

[0003] The purpose of this application is to provide a method for testing the heat dissipation performance of a radiator on a heat dissipation object, which can effectively reflect the actual heat dissipation effect that the radiator can play on the heat dissipation object.

[0004] The embodiments of this application are implemented as follows:

[0005] This application provides a method for testing the heat dissipation performance of a radiator on a heat dissipation object, comprising: applying a dye to a first preset surface, then attaching the radiator and the heat dissipation object according to the heat dissipation installation position; heating the attached radiator and the heat dissipation object to dry the dye; and separating the radiator and the heat dissipation object, then obtaining test indicators, and then judging the heat dissipation performance of the radiator on the heat dissipation object based on the test indicators.

[0006] The contact surface between the radiator and the object to be radiated is the first contact surface, and the contact surface between the object to be radiated and the radiator is the second contact surface. One of the first contact surface and the second contact surface is the first preset surface and the other is the second preset surface. The detection index is used to characterize the decolorization of the first preset surface and / or the staining of the second preset surface.

[0007] In the technical solution of this application, the radiator and the heat dissipation object are attached according to the heat dissipation installation position. By applying a dye to one of the contact surfaces, and after drying the dye and separating the radiator and the heat dissipation object, the discoloration of the contact surface with the dye and the staining of the contact surface without the dye can accurately reflect the contact between the radiator and the heat dissipation object at the installation position. This can effectively reflect the actual heat conduction performance and heat dissipation effect that the radiator can exert on the heat dissipation object.

[0008] In some possible implementations, obtaining detection indicators includes: acquiring an image of a target surface, which is a first preset surface and / or a second preset surface; and determining the detection indicators based on the image of the target surface.

[0009] In the above technical solution, the detection index is determined based on the image of the target surface, thereby avoiding the problem of manually measuring the area to evaluate the heat dissipation performance and effectively improving the detection accuracy of heat dissipation performance.

[0010] In some possible implementations, the detection metrics are determined based on the image of the target surface, including: identifying discolored and non-discolored areas in the image of the target surface; and determining the detection metrics based on the discolored and non-discolored areas.

[0011] In the above technical solution, the detection index is determined based on the color-changing and non-color-changing areas in the image of the target surface, thereby avoiding the problem of manually measuring the area to evaluate the heat dissipation performance and effectively improving the detection accuracy of heat dissipation performance.

[0012] In some possible implementations, identifying discolored and non-discolored regions in an image of the target surface includes: binarizing the image of the target surface to obtain a binarized image; and determining the discolored and non-discolored regions based on the pixel values ​​of the pixels in the binarized image.

[0013] In the above technical solution, the color-changing area and the non-color-changing area are determined based on the pixel values ​​of the pixels in the binarized image, thereby avoiding the problem of difficulty in determining the color-changing area and the non-color-changing area, and effectively improving the detection accuracy of heat dissipation performance.

[0014] In some possible implementations, determining the color-changed region and the uncolor-changed region based on the pixel values ​​of pixels in the binarized image includes: using a preset threshold to perform threshold segmentation on the pixel values ​​of pixels in the binarized image to obtain the color-changed region and the uncolor-changed region.

[0015] In the above technical solution, by using a preset threshold to perform threshold segmentation on the pixel values ​​of pixels in the binarized image, the problem of difficulty in determining the color-changing area and the non-color-changing area is avoided, and the detection accuracy of heat dissipation performance is effectively improved.

[0016] In some possible implementations, identifying discolored and non-discolored regions in an image of the target surface includes: using a neural network model to perform image semantic segmentation on the image of the target surface to obtain discolored and non-discolored regions.

[0017] In the above technical solution, by using a neural network model to perform image semantic segmentation on the target surface image to obtain the color-changing region and the non-color-changing region, the problem of difficulty in determining the adaptive threshold is avoided, and the accuracy of determining the color-changing region and the non-color-changing region based on the adaptive threshold is effectively improved.

[0018] In some possible implementations, the neural network model includes: a feature pyramid, an adaptive thresholding module, and a differentiable binarization module; image semantic segmentation of the target surface image using the neural network model includes: extracting features from the target surface image using the feature pyramid to obtain image features; performing differentiable binarization processing on the image features using the differentiable binarization module to obtain a differentiable binary image; and performing threshold segmentation on the differentiable binary image using the adaptive thresholding module to obtain color-changed regions and non-color-changed regions.

[0019] In the above technical solution, differentiable binarization processing is performed by using the differentiable binarization module in the neural network model, and threshold segmentation is performed on the differentiable binary image by using the adaptive threshold module. This avoids the problem of difficulty in determining the adaptive threshold and effectively improves the accuracy of determining the color-changing region and the non-color-changing region based on the adaptive threshold.

[0020] In some possible implementations, the detection metrics include: the ratio between the weight value of the color-changing region and the weight value of the non-color-changing region; determining the detection metrics based on the color-changing region and the non-color-changing region includes: for each pixel in the color-changing region, obtaining the pixel value and position coordinates of that pixel, determining the weight value of that pixel based on the pixel value and position coordinates, and obtaining the weight value of the color-changing region; for each pixel in the non-color-changing region, obtaining the pixel value and position coordinates of that pixel, determining the weight value of that pixel based on the pixel value and position coordinates, and obtaining the weight value of the non-color-changing region; dividing the weight value of the color-changing region by the weight value of the non-color-changing region to obtain the ratio between the weight value of the color-changing region and the weight value of the non-color-changing region.

[0021] In the above technical solution, when determining the detection index, not only the pixel value of the pixel is considered, but also the position coordinate of the pixel. This avoids the problem of determining the detection index based solely on the pixel value of the pixel, and effectively reflects the actual heat conduction performance and heat dissipation effect that the heat sink can exert on the heat dissipation object.

[0022] In some possible implementations, the detection metrics include at least one of the following: the ratio between the area of ​​the color-changing region and the area of ​​the non-color-changing region; or, the ratio between the number of pixels in the color-changing region and the number of pixels in the non-color-changing region; or, the ratio between the sum of the pixels in the color-changing region and the sum of the pixels in the non-color-changing region; or, the ratio between the weight value of the color-changing region and the weight value of the non-color-changing region, wherein the weight value is determined based on the pixel value and position coordinates of the pixel.

[0023] This application also provides an electronic device, including a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and when the machine-readable instructions are executed by the processor, perform the step of acquiring detection indicators as described in the possible implementations above.

[0024] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the step of acquiring detection indicators as described in the possible implementations above.

[0025] In some possible implementations, the dye is red ink.

[0026] In the above technical solution, red ink is used as the dyeing agent. Red ink has strong penetrability, making it easier for staining and discoloration to occur at the effective contact points between the radiator and the heat dissipation object. The bright color of red ink can also more clearly show the staining and discoloration phenomena.

[0027] In some possible implementations, the red ink comprises, by weight percentage: 30-50% butyl acid, 30-50% ethanol, 1-10% butanol, 1-5% nitrocellulose, 1-5% isopropanol, 1-5% n-propyl acetate, and 1-5% diacetone alcohol.

[0028] In the above technical solution, the red ink is formulated according to a specific composition ratio, so that the red ink has suitable properties such as viscosity, fluidity, permeability, adhesion, coverage, and fast drying. This ensures that the red ink can form a thin film with bright color, uniform thickness, and high coverage on the first preset surface through the coating process. After bonding and drying, the film formed by the red ink can also reliably exhibit coloring and decolorization phenomena, and is easy to observe.

[0029] In some possible implementations, the step of heating the heat sink and the heat dissipation object that are in contact with each other is carried out at a temperature of 80~100°C for a time of 3~5 hours.

[0030] In the above technical solution, the heat treatment is controlled according to a specific heating temperature and heating time to ensure that the dyeing agent is dried to a suitable degree, so that the dyeing and decolorization phenomena can more accurately reflect the contact between the radiator and the heat dissipation object at the installation position.

[0031] In some possible implementations, the heat dissipation target is the chip, the first preset surface is the first contact surface, and the detection index is the staining condition of the second preset surface.

[0032] In the above technical solution, the heat dissipation object is a chip. A dye is applied to the heat sink, and the heat dissipation performance of the heat sink is judged by the dyeing of the chip. Since the size of the dyed area on the second contact surface of the chip can show the effective contact area between the chip and the heat sink, the depth of the dye color can show the degree of contact between the chip and the heat sink, and the location of the dyed area can also show the basic position of the chip and the heat sink, it can more intuitively and comprehensively show the contact situation between the chip and the heat sink, thereby more intuitively and comprehensively reflecting the actual heat dissipation effect that the heat sink can play on the chip.

[0033] In some possible implementations, in the step of applying the dye to the first preset surface, the first preset surface is placed horizontally upwards.

[0034] In the above technical solution, placing the first preset surface horizontally upwards to apply the dye is beneficial for the dye to flow level and form a color film of uniform thickness, so that the dyeing and decolorization can more accurately reflect the contact between the radiator and the heat dissipation object at the installation position.

[0035] In some possible implementations, in the step of attaching the radiator and the heat dissipation object according to the heat dissipation installation position, the radiator and the heat dissipation object are brought close to each other in the heat dissipation installation direction to achieve attachment.

[0036] In the above technical solution, the radiator and the heat dissipation object are brought close to each other according to the heat dissipation installation direction to achieve contact, which simulates the installation process of the radiator and the heat dissipation object. This can more accurately simulate the contact situation of the radiator and the heat dissipation object, so that the staining and discoloration can more accurately reflect the contact between the radiator and the heat dissipation object at the installation position.

[0037] In some possible implementations, during the step of separating the radiator and the heat dissipation object, the radiator and the heat dissipation object move away from each other along a preset direction; wherein the preset direction is perpendicular to the first preset surface.

[0038] In the above technical solution, the radiator and the heat dissipation object are separated along a preset direction perpendicular to the first preset surface to avoid friction during the separation process of the radiator and the heat dissipation object, which would affect the decolorization and dyeing.

[0039] In some possible implementations, after the step of attaching the radiator and the heat dissipation object according to the heat dissipation installation position, and before the step of heating the attached radiator and the heat dissipation object, the method further includes: removing the dye located on the outer edge of the first contact surface and the second contact surface.

[0040] In the above technical solution, the dye on the outer edge of the first contact surface and the second contact surface is removed before heat treatment to avoid the dye on the outer edge interfering with the decolorization and dyeing of the contact surface.

[0041] In some possible implementations, prior to the step of applying the dye to the first preset surface, the method further includes cleaning the first and second contact surfaces to remove particulate contaminants and oil stains from the first and second contact surfaces.

[0042] In the above technical solution, the first and second contact surfaces are cleaned between the application of the dye to remove particulate contaminants and oil stains, so as to avoid the contaminants interfering with the uniform application of the dye and to avoid the contaminants interfering with the discoloration and staining phenomena that occur on the contact surfaces.

[0043] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the structure of an existing packaged chip.

[0045] Figure 2 A schematic diagram illustrating the process of determining detection indicators based on an image of a target surface, provided in an embodiment of this application;

[0046] Figure 3 A schematic diagram showing the color change of the heat sink and heat dissipation object before and after heat dissipation, provided in the embodiments of this application;

[0047] Figure 4 This is a schematic diagram of the network structure of the neural network model provided in the embodiments of this application;

[0048] Figure 5 This is a flowchart illustrating a method for detecting the heat dissipation performance of a heat sink on a heat dissipation object, as provided in an embodiment of this application. Attached Figure Description

[0049] Icons: 100 - Packaged chip; 110 - Chip; 120 - Chip solder; 130 - Upper copper layer; 140 - Ceramic; 150 - Lower copper layer; 160 - Substrate solder; 170 - Copper substrate.

[0050] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below. Where specific conditions are not specified in the embodiments, conventional conditions or conditions recommended by the manufacturer shall apply. Reagents or instruments whose manufacturers are not specified are all conventional products that can be purchased commercially.

[0051] It should be noted that the terms "and / or" in this application, such as "feature 1 and / or feature 2", all refer to the three cases of "feature 1" alone, "feature 2" alone, and "feature 1" plus "feature 2".

[0052] In addition, in the description of this application, unless otherwise stated, "one or more" means two or more; the range of "numerical value a to numerical value b" includes the two endpoints "a" and "b"; and "unit of measurement" in "numerical value a to numerical value b + unit of measurement" represents the "unit of measurement" of both "numerical value a" and "numerical value b".

[0053] The method for testing the heat dissipation performance of the radiator according to the embodiments of this application will be described in detail below.

[0054] With the rapid development of high-frequency, high-speed and integrated circuit technologies in electronic devices, the total power density of electronic components has increased significantly while their physical size has become smaller and smaller. The heat flux density has also increased accordingly. Therefore, high-temperature environments will inevitably affect the performance of electronic components, which requires more efficient thermal control.

[0055] Taking chips as an example, the most common packaging method for chips is currently modular packaging, such as... Figure 1 As shown, the packaged chip 100 of the module package includes a chip 110, a chip solder 120, an upper copper layer 130, a ceramic layer 140, a lower copper layer 150, a substrate solder 160, and a copper substrate 170 arranged sequentially. The upper copper layer 130, ceramic layer 140, and lower copper layer 150 constitute a DCB substrate (ceramic-based copper clad laminate). The top surface of the chip 110 is in contact with the bonding wires, and the bottom surface of the copper substrate 170 conducts heat to the heat sink via thermal grease. Detailed Implementation

[0056] Because current electronic chips operate under high temperature, high pressure, and high power environments, the limitations of traditional packaging heat dissipation, along with the high thermal conductivity and Young's modulus of silicon, force chips to withstand greater thermal stress during operation. Excessive thermal stress can lead to the detachment of internal bonding wires and the cracking of the solder layer, resulting in increased module thermal resistance, rising junction temperature, and ultimately device failure, potentially causing malfunctions in the entire electronic system. Therefore, chip temperature is a primary indicator of chip thermal performance, and chip heat dissipation is a prerequisite for chip thermal management, condition monitoring, and lifespan prediction. Therefore, it is crucial to determine whether a heat sink effectively conducts and transfers heat to the chip.

[0057] Currently, taking chips as an example, in the methods for testing the heat conduction and heat transfer performance of heat sinks on chips, a heating element simulating the heating of a microprocessor chip is usually used on the heat sink. Several thermocouples are evenly distributed on the heating element, and the heating power of the thermocouples is controlled by voltage. Then, the temperature of the heating element simulating the heating of a microprocessor chip is detected, thereby evaluating the heat dissipation performance of the heat sink on the chip.

[0058] The applicant's research found that when a gap is created between the chip and the heat sink due to poor contact, the gap between the chip and the heat sink significantly reduces the chip's heat dissipation performance, and may even prevent the heat sink from effectively dissipating heat from the chip, since air is a poor conductor of heat.

[0059] Current evaluation methods involve attaching a heating element that simulates the heat generated by a microprocessor chip to a heat sink, and then evaluating the heat sink's heat dissipation performance on the chip through heat control and temperature detection. However, this method can only evaluate the heat dissipation characteristics of the heat sink itself and cannot accurately reflect whether the heat sink can effectively contact the chip and play an actual heat dissipation role.

[0060] Based on this, the applicant discovered through in-depth research that by dyeing one of the contact surfaces between the heat sink and the chip, and then drying the dye after the heat sink and chip are bonded together, the effective contact area between the heat sink and the chip will experience discoloration and staining. By observing the discoloration and staining of the contact surface between the heat sink and the chip, the contact between the two can be directly reflected, thereby effectively reflecting the actual heat conduction performance and heat dissipation function that the heat sink can perform on the chip.

[0061] Based on the above research findings, this application provides a method for testing the heat dissipation performance of a radiator on a heat dissipation object, comprising: applying a dye to a first preset surface, then attaching the radiator and the heat dissipation object according to the heat dissipation installation position; heating the attached radiator and the heat dissipation object to dry the dye; separating the radiator and the heat dissipation object, then obtaining test indicators, and then judging the heat dissipation performance of the radiator on the heat dissipation object based on the test indicators.

[0062] The contact surface between the radiator and the object to be radiated is the first contact surface, and the contact surface between the object to be radiated and the radiator is the second contact surface. One of the first contact surface and the second contact surface is the first preset surface and the other is the second preset surface. The detection index is used to characterize the decolorization of the first preset surface and / or the staining of the second preset surface.

[0063] In this application, applying dye to the first preset surface means covering the first preset surface with dye, that is, the dye covers the entire first preset surface.

[0064] A heat dissipation object refers to an electronic device that comes into contact with a heat sink and dissipates heat through the heat sink, such as, but not limited to, a chip.

[0065] The installation positions of the radiator and the object being cooled correspond to the points where they are installed and in contact during actual operation. Specifically, the installation position of the radiator is the first contact surface, and the installation position of the object being cooled is the second contact surface.

[0066] Fitting the radiator and the object to be cooled together means fitting the radiator and the object to be cooled together as tightly as possible.

[0067] It should be noted that the dye applied to the first preset surface is dried during the heat treatment after the bonding process. Therefore, after the dye is applied to the first preset surface and the dye is evenly spread on the first preset surface, the dye is not dried, but the heat sink and the heat dissipation object are directly bonded together.

[0068] In the technical solution of this application, the radiator and the heat dissipation object are attached according to the heat dissipation installation position. By applying a dye to one of the contact surfaces, and after drying the dye and separating the radiator and the heat dissipation object, the discoloration of the contact surface with the dye and the staining of the contact surface without the dye can accurately reflect the contact between the radiator and the heat dissipation object at the installation position. This can effectively reflect the actual heat conduction performance and heat dissipation effect that the radiator can exert on the heat dissipation object.

[0069] In this application, the type of dye is not limited, as long as it can effectively cover the first preset surface during the application process and effectively dye the surface in contact with the first preset surface after the bonding and drying processes. As an example, the dye can be selected based on one or more of the following performance requirements: smooth flowability, strong penetration, fast drying, strong adhesion, thin film thickness, and clear and vibrant color. The dye can be a natural dye or a synthetic dye; synthetic dyes include, for example, red ink, blue ink, and black ink.

[0070] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating the process of determining detection indicators based on an image of a target surface, as provided in an embodiment of this application. As an optional implementation of the above-described method for detecting heat dissipation performance, the detection indicators can be determined based on an image of the target surface during the acquisition process. This implementation may include:

[0071] Step 210: Acquire an image of the target surface, which is a first preset surface and / or a second preset surface.

[0072] The target surface refers to the first preset surface and / or the second preset surface. There are many types of target surfaces, including but not limited to: the first type is the surface of the heat sink; the second type is the surface of the object to be heated; and the third type is the surface of both the heat sink and the object to be heated.

[0073] The image acquisition methods for the target surface in step 210 above include: First, using a terminal device such as a camera, video recorder, or color camera to capture an image of the target surface; then, the terminal device sends the image of the target surface to an electronic device, which receives the image and stores it in a file system, database, or removable storage device; Second, acquiring a pre-stored image of the target surface, specifically, for example, acquiring an image of the target surface from a file system, database, or removable storage device; Third, using software such as a browser to acquire an image of the target surface from the internet, or using other applications to access the internet to acquire an image of the target surface.

[0074] Step 220: Determine the detection index based on the image of the target surface.

[0075] As a first optional implementation of step 220 above, the detection index can be determined based on the color-changing and non-color-changing areas in the image. This implementation may include:

[0076] Step 221: Identify the discolored and undiscolored areas in the image of the target surface.

[0077] Please refer to Figure 3 , Figure 3 This is a schematic diagram showing the color change of the heat sink and the heat dissipation object before and after heat dissipation, as provided in the embodiments of this application. In practice, either the heat sink or the heat dissipation object can be dyed; for ease of description, this is not provided. Figure 3 The diagram only shows a schematic of dyeing a heat dissipation object. When dyeing a heat dissipation object, the discolored area is the same as the discolored area within the object (i.e.,...). Figure 3 The ring in the middle), and the area that did not change color is the area in the heat sink that did not discolor (i.e., the ring in the middle). Figure 3 (The area excluding the circular ring). Similarly, for a heatsink, the discolored area is the stained area within the heatsink (i.e., the area that is dyed). Figure 3 The ring in the middle), and the uncolored area is the undyed area in the heat sink (i.e., the ring in the middle). Figure 3 (The part excluding the ring).

[0078] Therefore, the aforementioned discolored and undiscolored areas can have many variations, including but not limited to: First, the discolored area is a decolorized area of ​​the first preset surface, and the undiscolored area is an undecolorized area of ​​the first preset surface; second, the discolored area is a stained area of ​​the second preset surface, and the undiscolored area is an unstained area of ​​the second preset surface. It can be understood that the first preset surface can be the surface of the radiator or the surface of the object being cooled. When the first preset surface is the surface of the radiator, the second preset surface is the surface of the object being cooled; similarly, when the first preset surface is the surface of the object being cooled, the second preset surface is the surface of the radiator.

[0079] It is understandable that there are many recognition methods, including but not limited to: threshold segmentation after binarization, image semantic segmentation of neural networks, etc. These recognition methods will be explained in detail below.

[0080] Step 222: Determine the detection indicators based on the discolored and non-discolored areas.

[0081] Since there are too many implementation schemes for steps 221 to 222, the implementation schemes for these two steps will be described in detail below.

[0082] As a first optional implementation of step 221 above, the implementation of identifying the color-changing areas and non-color-changing areas in the image may include:

[0083] Step 221a: Binarize the image of the target surface to obtain a binarized image.

[0084] An example implementation of step S221a above is as follows: For each pixel in the image of the target surface, it can be determined whether the pixel value is less than a preset threshold; if the pixel value is less than the preset threshold, the pixel value is set to zero; otherwise, the pixel value is set to one (or 255); after all pixels have been processed in the above way, a binarized image can be obtained. The preset threshold can be a manually specified threshold, such as 100, 200, or 250, etc. Alternatively, the preset threshold can be a globally adaptive threshold calculated using a neural network model or an executable program based on statistics.

[0085] Step 221b: Determine the color-changed and non-color-changed regions based on the pixel values ​​of the pixels in the binarized image.

[0086] In the first optional implementation of step 221b above, the color-changing area and the non-color-changing area of ​​the first preset surface can be determined based on the pixel value. For example, for each pixel in the binarized image, it is determined whether the pixel value of the pixel is greater than a preset threshold; if the pixel value of the pixel is greater than the preset threshold, the pixel is added to the non-color-changing area of ​​the first preset surface; if the pixel value of the pixel is less than the preset threshold, the pixel is added to the color-changing area of ​​the first preset surface.

[0087] In a second optional implementation of step 221b above, the color-changing area and the uncolor-changing area of ​​the second preset surface can be determined based on the pixel value. For example, for each pixel in the binarized image, it is determined whether the pixel value of the pixel is greater than a preset threshold; if the pixel value of the pixel is greater than the preset threshold, the pixel is added to the colored area of ​​the second preset surface; if the pixel value of the pixel is less than the preset threshold, the pixel is added to the uncolored area of ​​the second preset surface.

[0088] As a second optional implementation of step 221 above, identifying the discolored and non-discolored areas in the image of the target surface includes:

[0089] Step 221c: Use a neural network model to perform image semantic segmentation on the image of the target surface to obtain the discolored and undiscolored regions.

[0090] Please refer to Figure 4 , Figure 4 A schematic diagram of the network structure of the neural network model provided in this application embodiment; as an optional implementation of step 221c above, the neural network model may include: a feature pyramid, an adaptive thresholding module, and a differentiable binarization module; using the neural network model to perform image semantic segmentation on the target surface image includes:

[0091] Step 221d: Use the feature pyramid to extract features from the image of the target surface to obtain image features.

[0092] The aforementioned feature pyramid refers to a neural network stacked with neural networks of different sizes and / or different types. This feature pyramid is used for feature extraction and can be a feature extraction network stacked with convolutional networks and / or residual networks (ResNet).

[0093] An example implementation of step S221d above is as follows: A convolutional network is used to extract features from the image of the target surface to obtain image convolutional features. A residual network (ResNet) is then used to perform residual processing on the image convolutional features to obtain image features. The residual networks (ResNet) that can be used here include, but are not limited to, ResNet22, ResNet38, ResNet50, ResNet101, and ResNet152, etc.

[0094] Step 221e: Use the differentiable binarization module to perform differentiable binarization processing on the image features to obtain a differentiable binary image.

[0095] An example implementation of step 221e above involves using a differentiable binarization module that approximates a step function and is differentiable to perform differentiable binarization processing on image features to obtain a differentiable binary image. The aforementioned differentiable binarization module refers to a neural network module that uses an approximate binarization function that approximates a step function and is differentiable. Since the standard binarization function is not differentiable, this approximate binarization function is used instead of the standard binarization function. This function is called a differentiable binarization (DB) function, enabling the network to learn the text segmentation threshold end-to-end during training, thereby achieving a fully differentiable and automatically adaptive threshold adjustment effect.

[0096] Step 221f: Use the adaptive thresholding module to perform thresholding segmentation on the differentiable binary image to obtain the color-changed region and the uncolor-changed region.

[0097] An example implementation of step 221f above is as follows: An adaptive thresholding module is used to obtain the differentiable binary image calculated by the differentiable binarization module and an adaptive threshold. Then, the adaptive threshold is used to perform threshold segmentation on the differentiable binary image to obtain the color-changing regions and the uncolor-changing regions. Here, the aforementioned adaptive thresholding module refers to a neural network module capable of performing threshold segmentation on the differentiable binary image using an adaptive threshold. Specifically, it can employ network models such as LeNet, AlexNet, VGG, GoogLeNet, and ResNet, etc.

[0098] As a first optional implementation of step 222 above, the detection index may include: the ratio between the area of ​​the color-changing region and the area of ​​the non-color-changing region; the implementation of determining the detection index based on the color-changing region and the non-color-changing region may include: calculating the area of ​​the color-changing region in the image and calculating the area of ​​the non-color-changing region in the image, and then dividing the area of ​​the color-changing region by the area of ​​the non-color-changing region to obtain the ratio between the area of ​​the color-changing region and the area of ​​the non-color-changing region.

[0099] As a second optional implementation of step 222 above, the detection index may include: the ratio between the number of pixels in the color-changing region and the number of pixels in the non-color-changing region; the implementation of determining the detection index based on the color-changing region and the non-color-changing region may include: counting the number of pixels in the color-changing region of the image and counting the number of pixels in the non-color-changing region of the image, and then dividing the number of pixels in the color-changing region by the number of pixels in the non-color-changing region to obtain the ratio between the number of pixels in the color-changing region and the number of pixels in the non-color-changing region.

[0100] As a third optional implementation of step 222 above, the detection index may include: the ratio between the sum of pixels in the color-changing region and the sum of pixels in the non-color-changing region; the implementation of determining the detection index based on the color-changing region and the non-color-changing region may include: calculating the sum of pixels in the color-changing region of the image and calculating the sum of pixels in the non-color-changing region of the image, and then dividing the sum of pixels in the color-changing region by the sum of pixels in the non-color-changing region to obtain the ratio between the sum of pixels in the color-changing region and the sum of pixels in the non-color-changing region.

[0101] As a fourth optional implementation of step 222 above, the detection index may include: the ratio between the weight value of the color-changing area and the weight value of the non-color-changing area. It is understood that, due to the inherent limitations of the heat sink, for example, when a fan is dissipating heat, the airflow speed is higher near the fan blades and lower near the center point. Considering the different contributions of different position coordinates to the heat dissipation effect, the position coordinates of the pixel should be included to calculate its contribution weight to the heat dissipation effect (i.e., the weight value of that pixel). The above implementation of determining the detection index based on the color-changing and non-color-changing areas may include:

[0102] Step 222a: For each pixel in the color-changing region, obtain the pixel value and position coordinates of the pixel, determine the weight value of the pixel based on the pixel value and position coordinates, and obtain the weight value of the color-changing region.

[0103] An example implementation of step 222a above is: using formula W i =len(P i -C)×V i The pixel value and position coordinates of the pixel are used to calculate the weight value of the pixel; where W i P represents the weight value of the i-th pixel in the color-changing region. i Let represent the coordinates of the i-th pixel in the color-changing region, and C represent the coordinates of a preset pixel (e.g., the center point of a fan). This preset pixel can be set according to other scattering methods, such as water cooling. len(P i -C) represents the straight-line distance between the coordinates of the i-th pixel in the color-changing region and the coordinates of the preset pixel, V i This represents the pixel value of the i-th pixel in the color-changing region. If it is a value from the RGB three channels, the average of the RGB three channel values ​​can be used as the pixel value. The weight values ​​of each pixel in the color-changing region are summed to obtain the weight value of the color-changing region.

[0104] Step 222b: For each pixel in the uncolored region, obtain the pixel value and position coordinates of the pixel, determine the weight value of the pixel based on the pixel value and position coordinates, and obtain the weight value of the uncolored region.

[0105] Step 222c: Divide the weight value of the color-changed region by the weight value of the uncolor-changed region to obtain the ratio between the weight values ​​of the color-changed region and the uncolor-changed region.

[0106] In the above technical solution, when determining the detection index, not only the pixel value of the pixel is considered, but also the position coordinate of the pixel. This avoids the problem of determining the detection index based solely on the pixel value of the pixel, and effectively reflects the actual heat conduction performance and heat dissipation effect that the heat sink can exert on the heat dissipation object.

[0107] Based on the above, we can summarize that the aforementioned detection indicators are all data ratios between the color-changing area and the non-color-changing area. A larger ratio indicates a better heat dissipation effect of the heat sink on the object being cooled; conversely, a smaller ratio indicates a worse heat dissipation effect. The aforementioned detection indicators may include at least one of the following: Alternatively, the ratio between the weight value of the color-changing area and the weight value of the non-color-changing area; the ratio between the area of ​​the color-changing area and the area of ​​the non-color-changing area; or the ratio between the number of pixels in the color-changing area and the number of pixels in the non-color-changing area; or the ratio between the sum of pixels in the color-changing area and the sum of pixels in the non-color-changing area; or the ratio between the weight value of the color-changing area and the weight value of the non-color-changing area, where the weight value is determined based on the pixel value and position coordinates of the pixel. For the specific determination method, please refer to the fourth optional implementation scheme in step 222 above.

[0108] As a second optional implementation of step 220 above, and as a first optional implementation of step 220 above, the detection index can be determined based on the discoloration area and the target surface in the image. This implementation may include:

[0109] Step 223: Identify the discolored areas in the image of the target surface.

[0110] The discolored area mentioned above can be the decolorized area of ​​the first preset surface and / or the stained area of ​​the second preset surface.

[0111] The implementation principle and implementation scheme of step 223 are similar to those of step 221. Therefore, the implementation principle and implementation scheme will not be explained here. If there is anything unclear, please refer to the description of step 221.

[0112] Step 224: Determine the detection indicators based on the discoloration area and the target surface.

[0113] The aforementioned detection indicators include, but are not limited to: the ratio between the area of ​​the discolored region and the area of ​​the target surface; or, the ratio between the number of pixels in the discolored region and the number of pixels on the target surface; or, the ratio between the sum of the pixels in the discolored region and the sum of the pixels on the target surface; or, the ratio between the weight value of the discolored region and the weight value of the target surface, wherein the weight value can be determined based on the pixel value and position coordinates of the pixel. For the specific determination method, please refer to the fourth optional implementation scheme of step 222 above.

[0114] The implementation schemes for step 224 above include, but are not limited to, the following:

[0115] In a first optional implementation scheme, the aforementioned detection index may include: the ratio between the area of ​​the discolored region and the area of ​​the target surface; the implementation scheme for determining the detection index based on the discolored region and the target surface may include: calculating the area of ​​the discolored region in the image and calculating the area of ​​the target surface in the image, and then dividing the area of ​​the discolored region by the area of ​​the target surface to obtain the ratio between the area of ​​the discolored region and the area of ​​the target surface.

[0116] In a second alternative implementation, the detection index mentioned above may include the ratio between the number of pixels in the color-changing region and the number of pixels on the target surface. An implementation scheme for determining the detection index based on the color-changing region and the target surface may include: counting the number of pixels in the color-changing region of the image and counting the number of pixels on the target surface of the image; then, dividing the number of pixels in the color-changing region by the number of pixels on the target surface to obtain the ratio between the number of pixels in the color-changing region and the number of pixels on the target surface.

[0117] In a third alternative implementation, the aforementioned detection index may include the ratio between the sum of pixels in the color-changing region and the sum of pixels in the target surface. The implementation of determining the detection index based on the color-changing region and the target surface may include: calculating the sum of pixels in the color-changing region of the image and calculating the sum of pixels in the target surface of the image; then, dividing the sum of pixels in the color-changing region by the sum of pixels in the target surface to obtain the ratio between the sum of pixels in the color-changing region and the sum of pixels in the target surface.

[0118] The fourth optional implementation involves obtaining the pixel value and position coordinates of each pixel in the color-changing region, determining its weight value based on these coordinates, and thus obtaining the weight value of the color-changing region. Similarly, for each pixel on the target surface, the pixel value and position coordinates are obtained, and the weight value of the target surface is determined based on these coordinates, thus obtaining the weight value of the target surface. The weight value of the color-changing region is then divided by the weight value of the target surface to obtain the ratio between their respective weight values. It is understood that the calculation process for this ratio is similar in principle to the calculation processes in steps S222a to S222c above, and therefore will not be elaborated upon here.

[0119] It is understood that the above-mentioned test indicators are all data ratio values ​​between the discolored area and the target surface. A larger ratio value indicates a better heat dissipation effect of the radiator on the object being cooled, and conversely, a smaller ratio value indicates a worse heat dissipation effect. In practice, the correspondence between the above-mentioned test indicators and heat dissipation performance can be pre-calibrated or calculated in real time. Therefore, the above-mentioned implementation scheme for judging the heat dissipation performance of the radiator based on test indicators can include:

[0120] In the first implementation scheme, the correspondence between the test index and heat dissipation performance is pre-defined. For example, the implementation scheme for judging the heat dissipation performance of the radiator based on the test index is as follows: if the data ratio is within the range of 0% to 30%, the heat dissipation performance of the radiator is determined to be poor; if the data ratio is within the range of 30% to 60%, the heat dissipation performance is determined to be moderate; if the data ratio is within the range of 60% to 80%, the heat dissipation performance is determined to be good; and if the data ratio is within the range of 80% to 100%, the heat dissipation performance is determined to be excellent. The data ratios include: in the first case, the ratio between the discolored area and the undiscolored area; and in the second case, the ratio between the discolored area and the target surface.

[0121] In the second implementation scheme, the correspondence between the aforementioned test indicators and heat dissipation performance can also be calculated in real time. For example, in the implementation scheme of judging the heat dissipation performance of the radiator based on test indicators, since the correspondence between the test indicators and heat dissipation performance may be linear or non-linear, in practice, a linear or non-linear function can be used to pre-fit the correspondence between the test indicators and heat dissipation performance to obtain the corresponding function. After obtaining the data ratio value of the test indicator, the corresponding function can be used to map the data ratio value of the test indicator to obtain the corresponding heat dissipation performance. For example, the heat dissipation performance includes, but is not limited to: poor, medium, good, and excellent.

[0122] Studies have found that the type of dye has a significant impact on the occurrence and observation of staining and decolorization phenomena, and is one of the key factors in detection methods.

[0123] Among them, red ink has strong penetration and bright color. In some possible implementations, the dye is red ink, which makes it easier for the effective contact points between the heat sink and the heat dissipation object to be stained and discolored, and can also make the stained and discolored phenomena more obvious.

[0124] Further research revealed that the concentration of red ink affects its viscosity; excessively high concentration leads to poor flowability, while insufficient concentration results in a less pronounced color. Furthermore, the types of raw materials used in red ink influence its permeability, adhesion, coverage, and drying speed. Controlling the composition and formulation of red ink to achieve suitable concentration and raw material types ensures appropriate viscosity, flowability, permeability, adhesion, coverage, and drying speed. This guarantees that the red ink, when applied to a pre-defined surface, forms a vibrant, uniformly thick, and highly covert thin film, facilitating reliable subsequent dyeing and decolorization, and allowing for easy observation.

[0125] In some embodiments, the red ink is formulated according to a specific composition ratio, wherein the red ink comprises, by mass percentage: 30-50% butyl acetate (CAS No. 123-86-4), 30-50% ethanol (CAS No. 64-17-5), 1-10% n-butanol (CAS No. 71-36-3), 1-5% nitrocellulose (CAS No. 9004-70-0), 1-5% isopropyl alcohol (CAS No. 67-63-0), 1-5% propylacetate (CAS No. 109-60-4), and 1-5% diacetone alcohol (CAS No. 123-42-2).

[0126] The mass percentages of butyl acetate and ethanol are, for example, but not limited to, any one of 30%, 40%, and 50% or any range between them; the mass percentages of n-butanol are, for example, but not limited to, any one of 1%, 3%, 5%, 8%, and 10% or any range between them; and the mass percentages of nitrocellulose, isopropanol, n-propyl acetate, and diacetone alcohol are, for example, but not limited to, any one of 1%, 3%, and 5% or any range between them.

[0127] In this application, the red ink can be prepared according to the above composition ratio, or a ready-made red ink that meets the above composition ratio can be used directly. As an example, red ink of the American brand DYKEM and model DYKEMSTEELRED is selected.

[0128] The study also found that the degree of drying during heat treatment has a significant impact on the occurrence and observation of staining and decolorization phenomena, and is one of the key factors in the detection method. Controlling the heating time and temperature to achieve a suitable degree of drying during heat treatment helps ensure the accuracy of the detection method.

[0129] In some possible implementations, the step of heating the heat sink and the heat dissipation object that are in contact with each other is carried out at a temperature of 80 to 100°C for a duration of 3 to 5 hours.

[0130] The heating temperature is, for example, but not limited to, any one of 80℃, 85℃, 90℃, 95℃, and 100℃, or a range between any two. The heating time is, for example, but not limited to, any one of 3h, 3.5h, 4h, 4.5h, and 5h, or a range between any two. If the heating temperature is too low and / or the heating time is insufficient, resulting in incomplete drying, the dyeing agent will not be fully cured. When separating the radiator from the object being radiated, the undried dyeing agent may flow, thus affecting the observation and detection of the decolorized and stained areas. If the heating temperature is too high and / or the heating time is too long, resulting in over-drying, the dyeing agent will dry into a powdery state, making it impossible to accurately determine whether decolorization or staineding has occurred.

[0131] In some possible implementations, in the step of applying the dye to the first preset surface, the first preset surface is placed horizontally upwards. This design facilitates the leveling of the dye and the formation of a color film of uniform thickness.

[0132] In some possible implementations, in the step of attaching the radiator and the heat dissipation object according to the heat dissipation installation position, the radiator and the heat dissipation object are brought close to each other in the heat dissipation installation direction to achieve attachment. This scheme is designed to simulate the installation process of the radiator and the heat dissipation object, and can more accurately simulate the contact situation of the radiator and the heat dissipation object.

[0133] In some possible implementations, during the step of separating the radiator and the heat dissipation object, the radiator and the heat dissipation object move away from each other along a preset direction; wherein, the preset direction is perpendicular to the first preset surface. This design avoids friction during the separation of the radiator and the heat dissipation object, which could affect the decolorization and dyeing process.

[0134] Considering that there may be particulate contaminants and oil stains at the heat dissipation installation location of the radiator and heat dissipation object, the presence of contaminants will cause interference when applying dye and when the radiator and heat dissipation object are attached and undergo discoloration and dyeing reactions. Therefore, the heat dissipation installation location of the radiator and heat dissipation object can be cleaned first.

[0135] Based on the above considerations, in some possible implementations, before the step of applying the dye to the first preset surface, the method further includes cleaning the first contact surface and the second contact surface to remove particulate contaminants and oil stains from the first contact surface and the second contact surface.

[0136] Considering that some dye may be applied or flowed onto the outer edge of the first preset surface during the application of the dye, and that some dye may be squeezed onto the outer edge of the heat dissipation mounting position when the radiator and the heat dissipation object are attached, the slurry located on the outer edge of the heat dissipation mounting position may flow back to the first contact surface and / or the second contact surface during the subsequent drying and separation steps. This could interfere with the decolorization and dyeing process after the radiator and the heat dissipation object come into contact. Therefore, the dye located on the outer edge of the heat dissipation mounting position can be removed between the heat treatment steps to avoid interference.

[0137] Based on the above considerations, in some possible implementations, after the step of attaching the radiator and the heat dissipation object according to the heat dissipation installation position, and before the step of heating the attached radiator and the heat dissipation object, the method further includes: removing the dye located on the outer edge of the first contact surface and the second contact surface.

[0138] As a representative heat dissipation object, the chip is used in some possible implementations as the heat dissipation object, the first preset surface is the first contact surface, and the detection index is the staining condition of the second preset surface (which is the second contact surface). In this design, the size of the stained area on the second contact surface of the chip can show the effective contact area between the chip and the heat sink, the depth of the staining color can show the degree of contact between the chip and the heat sink, and the location of the stained area can also show the basic position of the chip and the heat sink. This allows for a more intuitive and comprehensive display of the contact condition between the chip and the heat sink, thus more intuitively and comprehensively reflecting the actual heat dissipation effect that the heat sink can exert on the chip.

[0139] See Figure 5 Using a chip as the heat dissipation target, in some exemplary embodiments, the method for testing the heat dissipation performance of the heat sink includes:

[0140] S1. Clean the sample.

[0141] (S1a) Determine the first contact surface of the heat sink and the second contact surface of the chip as the areas to be tested.

[0142] (S1b) Use an ultrasonic cleaner and isopropanol to clean the area to be tested to remove particulate contaminants and oil stains. Then, dry the sample with clean, oil-free cold air.

[0143] (S1c) Repeat step (S1b) once.

[0144] S2, Apply dye.

[0145] (S2a) Place the radiator horizontally on the platform so that the first contact surface of the radiator is facing upwards.

[0146] (S2b) Take red ink as a dye and apply it evenly to the first contact surface of the radiator.

[0147] S3, Adhere to the area to be tested

[0148] (S3a) According to the heat dissipation installation direction of the heat sink and the chip, bring the first contact surface of the heat sink coated with red ink and the second contact surface of the chip close to each other to achieve a tight fit.

[0149] S4. Cleaning treatment

[0150] (S4a) Remove the red ink from the outer edges of the first contact surface of the heat sink and the second contact surface of the chip.

[0151] S5, Heat Treatment

[0152] (S5a) Place the heat sink and chip that are attached to each other into an oven and dry them at a heating temperature of 90°C for 4 hours to dry the red ink. After drying, cool to room temperature.

[0153] S6. Result Acquisition and Judgment

[0154] (S6a) Wear protective gloves and use a clamp to separate the chip and the heat sink. During the separation process, avoid friction between the first contact surface and the second contact surface.

[0155] (S6b) The discoloration of the first contact surface of the heat sink and / or the staining of the second contact surface of the chip are observed under a microscope as test indicators. The heat sink's heat dissipation performance on the chip is determined based on these test indicators. The more severe the discoloration and staining, the denser the contact and the better the heat sink's heat dissipation performance on the chip.

[0156] This application also provides an electronic device, including a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the machine-readable instructions, when executed by the processor, perform the method as described in steps 210 to 220 above.

[0157] This application also provides a computer-readable storage medium storing a computer program, which is executed by a processor to perform the methods described in steps 210 to 220 above.

[0158] The computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0159] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0160] The embodiments described above are some, but not all, of the embodiments of this application. The detailed description of the embodiments of this application is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

Claims

1. A method for testing the heat dissipation performance of a radiator on a heat dissipation object, characterized in that, include: A dye is applied to a first preset surface, and then the radiator and the heat dissipation object are attached according to the heat dissipation installation position; wherein, the contact surface between the radiator and the heat dissipation object is the first contact surface, the contact surface between the heat dissipation object and the radiator is the second contact surface, and one of the first contact surface and the second contact surface is the first preset surface and the other is the second preset surface; The heat sink and the heat dissipation object, which are attached to each other, are subjected to heat treatment to dry the dyeing agent; and The heat sink and the heat dissipation object are separated, an image of the target surface is obtained, and a detection index is determined based on the image of the target surface. The heat dissipation performance of the heat sink on the heat dissipation object is judged based on the detection index. The target surface is the first preset surface and / or the second preset surface, and the detection index is used to characterize the decolorization of the first preset surface and / or the staining of the second preset surface.

2. The method according to claim 1, characterized in that, Determining the detection index based on the image of the target surface includes: Identify the discolored and undiscolored areas in the image of the target surface; The detection index is determined based on the color-changing area and the non-color-changing area.

3. The method according to claim 2, characterized in that, The identification of discolored and non-discolored areas in the image of the target surface includes: The image of the target surface is binarized to obtain a binarized image; The color-changing region and the uncolor-changing region are determined based on the pixel values ​​of the pixels in the binarized image.

4. The method according to claim 3, characterized in that, Determining the color-changed region and the uncolor-changed region based on the pixel values ​​of pixels in the binarized image includes: A preset threshold is used to perform threshold segmentation on the pixel values ​​of the pixels in the binarized image to obtain the color-changed region and the uncolor-changed region.

5. The method according to claim 2, characterized in that, The identification of discolored and non-discolored areas in the image of the target surface includes: The target surface image is semantically segmented using a neural network model to obtain the discolored region and the undiscolored region.

6. The method according to claim 5, characterized in that, The neural network model includes: a feature pyramid, an adaptive thresholding module, and a differentiable binarization module; the step of using the neural network model to perform image semantic segmentation on the target surface image includes: The feature pyramid is used to extract features from the image of the target surface to obtain image features; The image features are processed using the differentiable binarization module to obtain a differentiable binary image; The adaptive thresholding module is used to perform thresholding on the differentiable binary image to obtain the color-changing region and the uncolor-changing region.

7. The method according to claim 2, characterized in that, The detection index includes: the ratio between the weight value of the color-changing region and the weight value of the non-color-changing region; determining the detection index based on the color-changing region and the non-color-changing region includes: For each pixel in the color-changing region, obtain the pixel value and position coordinates of the pixel, determine the weight value of the pixel based on the pixel value and position coordinates, and obtain the weight value of the color-changing region. For each pixel in the uncolored region, obtain the pixel value and position coordinates of the pixel, determine the weight value of the pixel based on the pixel value and position coordinates, and obtain the weight value of the uncolored region. Divide the weight value of the color-changing region by the weight value of the uncolor-changing region to obtain the ratio between the weight value of the color-changing region and the weight value of the uncolor-changing region.

8. The method according to any one of claims 2-6, characterized in that, The detection indicators include at least one of the following: The ratio between the area of ​​the discolored region and the area of ​​the undiscolored region; Alternatively, the ratio between the number of pixels in the color-changing region and the number of pixels in the uncolor-changing region; Alternatively, the ratio between the sum of pixels in the color-changing region and the sum of pixels in the uncolor-changing region; Alternatively, it can be the ratio between the weight value of the color-changing region and the weight value of the uncolor-changing region, wherein the weight value is determined based on the pixel value and position coordinates of the pixel.

Images