Thermographic image optimization method and apparatus

Abstract

A method and device for optimizing thermal imaging images, the method comprising: collecting, by a data collection circuit, a raw image from a thermal imaging device in a detection field; updating, by a processor, a neural network model using a plurality of sample images stored in a storage and a plurality of sample correspondence tables corresponding to the plurality of sample images, wherein the plurality of sample correspondence tables indicate a correspondence between pixel values in the plurality of sample images generated by the thermal imaging device in a plurality of training fields and pixel values in optimized sample images; inputting, by the processor, the raw image into the neural network model to generate an optimized correspondence table; and converting, by the processor, the raw image into an optimized image using the optimized correspondence table. Thus, the problem of the complicated process of manually adjusting parameters and the problem of wasting time to adjust parameters for different fields in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to a thermal imaging processing technology, and more particularly to a thermal imaging image optimization method and apparatus. Background Technology

[0002] Due to the diversity of targets detected by thermal imagers, varying detection purposes, and differing individual subjective perceptions, it has been difficult to establish a standardized adjustment process for parameters such as contrast and brightness in previous thermal imaging technologies. Consequently, conventional thermal imaging techniques often required users to perform complex manual parameter adjustments to obtain optimized thermal images (i.e., reducing the influence of the environment or surrounding objects on the detected temperature of each target). Furthermore, previous thermal imaging techniques also required users to readjust parameters for different environments to obtain parameters suitable for the current context. Therefore, how to avoid the cumbersome process of manual parameter adjustment and the time-consuming readjustment required for various environments is a problem that those skilled in the art urgently need to solve. Summary of the Invention

[0003] The purpose of this invention is to provide a thermal imaging image optimization method and apparatus, which solves the problems of the complicated process of manually adjusting parameters and the time-consuming readjustment of parameters for various fields required by previous technologies.

[0004] To achieve the above objectives, this invention proposes an image optimization method for thermal imaging, comprising: Step a) Acquire a raw image from a thermal imaging device in a detection field through a data acquisition circuit; Step b) A neural network model is updated by a processor using a plurality of sample images stored in a memory and a plurality of sample correspondence tables corresponding to the plurality of sample images respectively, wherein the plurality of sample correspondence tables indicate the correspondence between the pixel values ​​in the plurality of sample images generated by the thermal imaging device in a plurality of training fields and the pixel values ​​in the optimized plurality of sample images. Step c) The processor inputs the original image into the neural network model to generate an optimized correspondence table; and Step d) The processor uses the optimized mapping table to convert the original image into an optimized image.

[0005] In one embodiment, the optimized correspondence table indicates the correspondence between multiple pixel value ranges of the original image and multiple optimized pixel values, wherein each sample correspondence table includes multiple pixel value ranges of the sample image corresponding to each sample correspondence table and multiple optimized pixel values ​​respectively corresponding to the multiple pixel value ranges.

[0006] In one embodiment, step b) includes: The processor uses these multiple sample images as multiple training samples. The processor uses the multiple sample correspondence tables corresponding to the multiple sample images as multiple training labels corresponding to the multiple training samples; and The processor updates the neural network model using the multiple training samples and the multiple training labels corresponding to the multiple training samples.

[0007] In one embodiment, step d) includes: using the processor, converting the pixel values ​​of the plurality of pixel coordinates of the original image into the pixel values ​​of the plurality of pixel coordinates of the optimized image via the optimized mapping table in a lookup manner.

[0008] In one embodiment, the optimization correspondence table includes multiple pixel value ranges of the original image and multiple optimized pixel values ​​corresponding to each of the multiple pixel value ranges, wherein step d) includes: The processor identifies which pixel value range each pixel coordinate belongs to from the optimization mapping table, and retrieves the optimized pixel value corresponding to each pixel value range from the optimization mapping table; and The processor sets the pixel value of each pixel coordinate in the optimized image to the corresponding optimized pixel value.

[0009] To achieve the above objectives, the present invention proposes an image optimization device for thermal imaging, comprising: A data acquisition circuit is configured to acquire a raw image from a thermal imaging device in a detection field; A storage device configured to store multiple sample images, multiple samples corresponding to each of the multiple sample images, and multiple instructions, wherein the multiple sample correspondence table indicates the correspondence between pixel values ​​in the multiple sample images generated by the thermal imaging device in multiple training fields and pixel values ​​in the optimized multiple sample images; and A processor, connected to the data acquisition circuit and the memory, is configured to run a neural network model and access the plurality of instructions to perform the following steps: Step a) Update the neural network model using the multiple sample images and the multiple sample correspondence table; Step b) Input the original image into the neural network model to generate an optimized correspondence table; and Step c) Use the optimized mapping table to convert the original image into an optimized image.

[0010] In one embodiment, the optimized correspondence table indicates the correspondence between multiple pixel value ranges of the original image and multiple optimized pixel values, wherein each sample correspondence table includes multiple pixel value ranges of the sample image corresponding to each sample correspondence table and multiple optimized pixel values ​​respectively corresponding to the multiple pixel value ranges.

[0011] In one embodiment, in action a), the processor is configured to perform the following action: Use these multiple sample images as multiple training samples; The multiple sample correspondence tables corresponding to the multiple sample images are used as multiple training labels corresponding to the multiple training samples respectively; and The neural network model is updated using the multiple training samples and the multiple training labels corresponding to the multiple training samples.

[0012] In one embodiment, in action c), the processor is configured to perform the following action: converting the pixel values ​​of the plurality of pixel coordinates of the original image into the pixel values ​​of the plurality of pixel coordinates of the optimized image by means of a lookup table via the optimized mapping table.

[0013] In one embodiment, the optimization mapping table includes multiple pixel value ranges of the original image and multiple optimized pixel values ​​corresponding to each of the multiple pixel value ranges, wherein in action c), the processor is configured to perform the following actions: Identify which pixel value range each pixel coordinate belongs to from the optimization correspondence table, and retrieve the optimized pixel value corresponding to each pixel value range from the optimization correspondence table; and The pixel value of each pixel coordinate in the optimized image is set to the corresponding optimized pixel value.

[0014] Compared to previous technologies, this invention utilizes a large number of pre-stored sample images and a sample correspondence table corresponding to specific parameters to train a neural network model. The trained neural network model then converts new images into a new correspondence table, and subsequently optimizes all pixels of the new image using a lookup table. This avoids the cumbersome process of manually adjusting parameters and the time-consuming readjustment required for various fields. Attached Figure Description

[0015] Figure 1 This is a block diagram of a thermal imaging image optimization device 100 in some embodiments of the present invention.

[0016] Figure 2 This is a schematic diagram of a sample correspondence table and sample images in some embodiments of the present invention.

[0017] Figure 3This is a flowchart of an image optimization method for thermal imaging in some embodiments of the present invention.

[0018] Figure 4 This is a schematic diagram of the optimized image in some embodiments of the present invention.

[0019] In the picture: 100: Image optimization device for thermal imaging; 110: Data acquisition circuit; 120: Storage device; 130: Processor; 131: Neural network model; si1~siN: Sample images; st1~stN: Sample mapping table; img1: Original image; int: Maximum pixel value range; osi1: Optimize sample images; S310~S340: Steps; ot1: Optimize the corresponding table; img2: Optimize image. Detailed Implementation

[0020] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention. (Refer to...) Figure 1 , Figure 1 A block diagram illustrating a thermal imaging image optimization apparatus 100 according to some embodiments of the present invention is shown. Figure 1 As shown, in this embodiment, the thermal imaging image optimization device 100 includes a data acquisition circuit 110, a storage device 120, and a processor 130. The processor 130 is coupled to the data acquisition circuit 110 and the storage device 120.

[0021] In some embodiments, the thermal imaging image optimization device 100 can be established by any data processing device (e.g., a desktop computer, laptop computer, or tablet computer) or server (e.g., a cloud server, virtual server, or rack server). In this embodiment, the data acquisition circuit 110 is used to acquire the original image img1 from the thermal imaging device 200 in the detection field. In other words, the user can use the thermal imaging device 200 to photograph the detection field to generate the original image img1 of the detection field. Then, the thermal imaging image optimization device 100 can be connected to the thermal imaging device 200 through the data acquisition circuit 110 to acquire the original image img1 of the detection field from the thermal imaging device 200.

[0022] In some embodiments, the detection field is the field captured by the thermal imaging device 200 to generate the original image img1 (e.g., street, factory, station, etc.). In some embodiments, the thermal imaging device 200 can be implemented by any thermal imager (e.g., a general infrared thermal imager, a quantum imager, or an optical and infrared composite imager, etc.). In some embodiments, the original image img1 can be an unoptimized image of any type (e.g., a grayscale image or an RGB image, etc.) generated by the thermal imaging device 200. In some embodiments, the data acquisition circuit 110 can be any wireless communication circuit (e.g., a Wi-Fi communication circuit or a Bluetooth communication circuit) or a wired communication circuit (e.g., an Ethernet communication circuit) used for communication.

[0023] In this embodiment, the storage device 120 is used to store multiple sample images si1~siN, multiple mapping tables st1~stN corresponding to the sample images si1~siN, and multiple instructions, where N can be any positive integer without particular limitation. In some embodiments, the sample images si1~siN are multiple unoptimized images (i.e., the pixel values ​​of all pixels in the images are unprocessed thermal data) captured in advance by the thermal imaging device 200 in multiple training fields, and the sample images si1~siN and the original image img1 are images of the same type (e.g., both are grayscale images), where the training fields can be the same as or different from the detection fields. In some embodiments, the multiple instructions can be implemented by any firmware or software, and the processor 130 can access these instructions to execute the thermal imaging image optimization method described in subsequent paragraphs. In some embodiments, the storage device 120 can be implemented by flash memory, read-only memory, hard disk, or any equivalent storage component.

[0024] In this embodiment, the sample correspondence tables st1~stN indicate the correspondence between pixel values ​​in sample images si1~siN generated by the thermal imaging device 200 in multiple training fields and pixel values ​​in optimized sample images si1~siN. In some embodiments, the sample correspondence tables st1~stN can be multiple correspondence tables generated based on the sample images si1~siN using any image optimization algorithm (e.g., histogram equalization algorithm) that optimizes for contrast and brightness, etc. Specifically, the user can pre-set the optimal contrast and brightness parameters of a specific image optimization algorithm for the sample images si1~siN captured in multiple training fields, and the sample images si1~siN can be converted into sample correspondence tables st1~stN respectively by a specific image optimization algorithm (i.e., having a one-to-one conversion relationship).

[0025] The following explanation uses sample correspondence table st1 and sample image si1 as examples. Please refer to [the documentation / reference]. Figure 2 , Figure 2 A schematic diagram illustrating the sample correspondence table st1 and sample image si1 in some embodiments of the present invention is shown. Figure 2 As shown, the sample image si1 can be pre-converted into a sample correspondence table st1 by a histogram equalization algorithm (for example, statistical histograms of multiple pixel value intervals are generated by statistically analyzing the pixel values ​​of all pixels in the sample image si1, and then the sample correspondence table st1 is generated from the equalized statistical histograms). The sample correspondence table st1 includes multiple pixel value intervals of all pixels in the sample image si1 and the optimized pixel values ​​corresponding to each pixel value interval. The degree of histogram equalization can be adjusted by the user by setting optimal alpha parameters for multiple training fields in advance. In this way, the pixel values ​​of all pixels in the sample image si1 can be converted into their respective optimized pixel values ​​through the sample correspondence table st1.

[0026] In detail, as shown in the sample correspondence table st1, when the pixel value of one pixel coordinate in the sample image si1 (e.g., (10, 20)) falls within the range of 0 to 5, the corresponding optimized pixel value is 6. The pixel value of the same pixel coordinate in the optimized sample image osi1 (e.g., (10, 20)) can also be set to 6. When the pixel value of another pixel coordinate in the sample image si1 (e.g., (50, 10)) falls within the range of 6 to 10, the corresponding optimized pixel value is 8. The pixel value of the same pixel coordinate in the optimized sample image osi1 (e.g., (50, 10)) can also be set to 8. And so on, the pixel values ​​of all pixels in the sample image si1 can be converted to the pixel values ​​of all pixels in the optimized sample image osi1 via the sample correspondence table st1 using a lookup table. Other sample correspondence tables st2~stN also contain data similar to sample correspondence table st1, and are used to convert the pixel values ​​of all pixels in each of the sample images si2~siN into the pixel values ​​of all pixels in their respective optimized sample images. It is worth noting that the maximum pixel value range int of sample correspondence table st1 can be determined by the data size of each of the stored sample images si1~siN (for example, if the data size is 256 bits, the maximum pixel value range int can be a pixel value range of 250~255; while if the data size is 16384 bits, the maximum pixel value range int can be a pixel value range of 16380~16383).

[0027] Back Figure 1In this embodiment, the processor 130 further executes a neural network model 131. In some embodiments, the neural network model 131 can be implemented by any neural network model for image processing (e.g., a convolutional neural network model, a deep neural network model, a YOLO model, or a transformer model, etc.). In some embodiments, the processor 130 can be implemented by a central processing unit (CPU), a microcontroller unit (MCU), a programmable logic controller (PLC), a system-on-a-chip (SoC), or a field-programmable gate array (FPGA), etc.

[0028] Refer to together Figure 3 , Figure 3 A flowchart illustrating an image optimization method for thermal imaging in some embodiments of the present invention is shown. This image optimization method for thermal imaging is applicable to... Figure 1 The thermal imaging image optimization device 100 shown is illustrated. Figure 3 As shown, the thermal imaging image optimization method includes steps S310 to S340.

[0029] First, in step S310, the data acquisition circuit 110 acquires the original image img1 from the thermal imaging device 200 in the detection field. In step S320, the processor 130 updates the neural network model 131 using the sample images si1~siN stored in the storage 120 and the sample correspondence tables st1~stN.

[0030] In some embodiments, the processor 130 uses each sample image as a training sample and the sample correspondence table corresponding to each sample image as a training label. Next, the processor 130 inputs each training sample into the neural network model 131 to generate a corresponding result correspondence table as the result label, and calculates the loss value between the corresponding result label and the corresponding training label. Then, the processor 130 uses the calculated loss value to perform a backpropagation algorithm on the neural network model 131 to update the parameters in the neural network model 131 (i.e., the weight values ​​of each of the multiple neural network layers in the neural network model 131). In this way, the processor 130 can complete the above-mentioned update of the neural network model 131 (i.e., complete the training phase).

[0031] In some embodiments, the processor 130 may first perform any type of normalization on the sample images si1~siN (e.g., min-max normalization or Z-score normalization) to generate normalized sample images si1~siN, and then use each normalized sample image as a training sample.

[0032] In step S330, the processor 130 inputs the original image img1 into the neural network model 131 to generate an optimized correspondence table. In other words, whenever the thermal imaging device 200 captures a detection field to generate an original image img1, the processor 130 can use this original image img1 as input to the neural network model 131 to convert the original image img1 into an optimized correspondence table (i.e., in the usage stage). In some embodiments, the optimized correspondence table indicates the correspondence between pixel values ​​in the original image img1 and optimized pixel values. Specifically, when the pixel value of a pixel in the original image img1 belongs to a pixel value range, the optimized correspondence table can indicate the optimized pixel value corresponding to this pixel value range. In some embodiments, the processor 130 may also first perform any type of normalization on the original image img1 (e.g., min-max normalization or Z-score normalization, etc.) to generate a normalized original image img1, and then use the normalized original image img1 as input to the neural network model 131.

[0033] It is worth noting that the trained neural network model 131 can automatically adjust various parameters (i.e., brightness and contrast, etc.) corresponding to the original image img1 for any detection field to generate an optimized correspondence table. Therefore, this method eliminates the need for manual adjustment of various parameters to generate an optimized correspondence table to the original image img1.

[0034] In step S340, the processor 130 converts the original image img1 into an optimized image using an optimized mapping table. In some embodiments, the processor 130 converts the pixel values ​​of multiple pixel coordinates of the original image img1 into optimized pixel values ​​by looking up a table in the optimized mapping table, thereby converting the original image img1 into an optimized image. In some embodiments, the processor 130 identifies which pixel value range each pixel value belongs to from the optimized mapping table, and obtains the optimized pixel value corresponding to each pixel value range from the optimized mapping table. Then, the processor 130 sets each pixel coordinate in the original image img1 to the corresponding optimized pixel value, thereby converting the original image img1 into an optimized image. In some embodiments, the optimized mapping table indicates the correspondence between multiple pixel value ranges of the original image and the optimized pixel values.

[0035] Specifically, when the pixel value of a pixel in the original image img1 falls within a certain pixel value range, the processor 130 can look up the optimization correspondence table to obtain the optimized pixel value corresponding to this pixel value range, and then convert the pixel value of this pixel in the original image img1 into the corresponding optimized pixel value. Similarly, the processor 130 can convert the pixel values ​​of other pixels in the original image img1 in the same way. In this way, the processor 130 can convert the original image img1 into an optimized image. In some embodiments, the optimized image is of the same type as the original image img1 (for example, both the optimized image and the original image img1 are grayscale images).

[0036] The following example illustrates the generation of optimized images. (See also...) Figure 4 , Figure 4 A schematic diagram illustrating the optimized image img2 in some embodiments of the present invention is shown. For example... Figure 4 As shown, the processor 130 can input the original image img1 into the neural network model 131 to generate an optimized correspondence table ot1. The optimized correspondence table ot1 includes multiple pixel value ranges of all pixels in the original image img1 and optimized pixel values ​​corresponding to each pixel value range.

[0037] Next, by querying the optimization correspondence table ot1, the processor 130 can identify which pixel value range in the optimization correspondence table ot1 each pixel in the original image img1 belongs to. Then, the processor 130 converts the pixel values ​​of pixels in the original image img1 belonging to the 0-5 range to 3, and converts the pixel values ​​of pixels in the original image img1 belonging to the 6-10 range to 9. And so on, the processor 130 can convert the pixel values ​​of all pixels in the original image img1 to their respective optimized pixel values ​​by looking up the table in the optimization correspondence table ot1.

[0038] In other words, when the pixel value of one pixel coordinate (e.g., (11, 21)) in the original image img1 falls within the range of 0 to 5, the processor 130 can identify the corresponding optimized pixel value as 3 by looking up the optimization correspondence table ot1. The processor 130 can then set the pixel value of the same pixel coordinate (e.g., (11, 21)) in the optimized image img2 to 3. When the pixel value of another pixel coordinate (e.g., (55, 15)) in the original image img1 falls within the range of 6 to 10, the processor 130 can identify the corresponding optimized pixel value as 9 by looking up the optimization correspondence table ot1. The processor 130 can then set the pixel value of the same pixel coordinate (e.g., (55, 15)) in the optimized image img2 to 9.

[0039] Similarly, the processor 130 can look up the pixel values ​​of all pixels in the original image img1 and convert them into the pixel values ​​of all pixels in the optimized image img2 via the optimized correspondence table ot1. In this way, whenever the thermal imaging device 200 captures a new detection area to generate a new original image, the processor 130 can convert the new original image into a new optimized correspondence table and quickly convert the new original image into a new optimized image via the new optimized correspondence table, without requiring the manual parameter adjustment process of previous technologies for different target object diversity, detection purposes, personal subjective perception, and the shooting area. Furthermore, the method of optimizing pixels using the generated optimized correspondence table can achieve similar optimization effects to the sample correspondence tables st1~stN (i.e., similar contrast and brightness optimization). It is worth noting that the maximum pixel value range int of the optimized corresponding table ot1 can be determined by the data size of the original image img1 (for example, if the data size is 256 bits, the maximum pixel value range int can be a pixel value range of 250~255; while if the data size is 16384 bits, the maximum pixel value range int can be a pixel value range of 16380~16383).

[0040] In some embodiments, the thermal imaging image optimization device 100 may further include a display (not shown), and the processor 130 may control the display to show the optimized image img2, so that the user can view the optimized image img2 after optimization by the neural network model 131. It is worth noting that the displayed optimized image img2 has optimal contrast and brightness, etc. Furthermore, the optimization characteristics of the optimization correspondence table ot1 generated by the neural network model 131 can be similar to the optimization characteristics of the sample correspondence tables st1~stN (that is, parameters such as contrast and brightness can be adjusted similarly).

[0041] In summary, the thermal imaging image optimization method and apparatus proposed in this invention can train a neural network model using a large number of pre-stored sample images and a sample correspondence table corresponding to specific parameters. Therefore, whenever the thermal imaging device captures a new image, the proposed thermal imaging image optimization method and apparatus can use the trained neural network model to convert the new image into a new correspondence table, and then optimize all pixels of the new image (i.e., obtain the optimal contrast and brightness) by looking up the table. This avoids the cumbersome process of manually adjusting parameters and the time-consuming readjustment required for various environments. Furthermore, it achieves image optimization and maintains optimal optimization results without the need for complex image optimization algorithms.

[0042] The embodiments described above are merely preferred embodiments for fully illustrating the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention.

Claims

1. A thermal imaging image optimization method, characterized in that, include: Step a) Acquire a raw image from a thermal imaging device in a detection field through a data acquisition circuit; Step b) A neural network model is updated by a processor using a plurality of sample images stored in a memory and a plurality of sample correspondence tables corresponding to the plurality of sample images respectively, wherein the plurality of sample correspondence tables indicate the correspondence between the pixel values ​​in the plurality of sample images generated by the thermal imaging device in a plurality of training fields and the pixel values ​​in the optimized plurality of sample images. Step c) The original image is input into the neural network model through the processor to generate an optimized correspondence table; as well as Step d) The processor uses the optimized mapping table to convert the original image into an optimized image.

2. The thermal imaging image optimization method as described in claim 1, characterized in that, The optimized correspondence table indicates the correspondence between multiple pixel value ranges of the original image and multiple optimized pixel values. Each sample correspondence table includes multiple pixel value ranges of the sample image corresponding to each sample correspondence table and multiple optimized pixel values ​​corresponding to each of the multiple pixel value ranges.

3. The thermal imaging image optimization method as described in claim 1, characterized in that, Step b) includes: The processor uses these multiple sample images as multiple training samples. The processor uses the multiple sample correspondence tables corresponding to the multiple sample images as multiple training labels corresponding to the multiple training samples; and The processor updates the neural network model using the multiple training samples and the multiple training labels corresponding to the multiple training samples.

4. The thermal imaging image optimization method as described in claim 1, characterized in that, Step d) includes: The processor uses a lookup table to convert the pixel values ​​of multiple pixel coordinates of the original image into the pixel values ​​of the multiple pixel coordinates of the optimized image.

5. The thermal imaging image optimization method as described in claim 4, characterized in that, The optimization correspondence table includes multiple pixel value ranges of the original image and multiple optimized pixel values ​​corresponding to each of the multiple pixel value ranges. Step d) includes: The processor identifies which pixel value range each pixel coordinate belongs to from the optimization mapping table, and retrieves the optimized pixel value corresponding to each pixel value range from the optimization mapping table; and The processor sets the pixel value of each pixel coordinate in the optimized image to the corresponding optimized pixel value.

6. An image optimization device for thermal imaging, characterized in that, include: A data acquisition circuit is configured to acquire a raw image from a thermal imaging device in a detection field; A storage device is configured to store multiple sample images, multiple sample correspondence tables corresponding to the multiple sample images, and multiple instructions, wherein the multiple sample correspondence tables indicate the correspondence between pixel values ​​in the multiple sample images generated by the thermal imaging device in multiple training fields and pixel values ​​in the optimized multiple sample images. as well as A processor, connected to the data acquisition circuit and the memory, is configured to run a neural network model and access a plurality of instructions to perform the following actions: Action a) Update the neural network model using the multiple sample images and the multiple sample correspondence table; Action b) Input the original image into the neural network model to generate an optimized correspondence table; and Action c) Use the optimized mapping table to convert the original image into an optimized image.

7. The thermal imaging image optimization apparatus as described in claim 6, characterized in that, The optimized correspondence table indicates the correspondence between multiple pixel value ranges of the original image and multiple optimized pixel values. Each sample correspondence table includes multiple pixel value ranges of the sample image corresponding to each sample correspondence table and multiple optimized pixel values ​​corresponding to each of the multiple pixel value ranges.

8. The thermal imaging image optimization apparatus as described in claim 6, characterized in that, In action a), the processor is configured to perform the following action: Use these multiple sample images as multiple training samples; The multiple sample correspondence tables corresponding to the multiple sample images are used as multiple training labels corresponding to the multiple training samples respectively; as well as The neural network model is updated using the multiple training samples and the multiple training labels corresponding to the multiple training samples.

9. The thermal imaging image optimization apparatus as described in claim 6, characterized in that, In action c), the processor is configured to perform the following action: The pixel values ​​of multiple pixel coordinates of the original image are converted into the pixel values ​​of the same multiple pixel coordinates of the optimized image by means of a lookup table through the optimized mapping table.

10. The thermal imaging image optimization apparatus as described in claim 9, characterized in that, The optimization mapping table includes multiple pixel value ranges of the original image and multiple optimized pixel values ​​corresponding to each of the multiple pixel value ranges. In action c), the processor is configured to perform the following actions: Identify which pixel value range each pixel coordinate belongs to from the optimization correspondence table, and retrieve the optimized pixel value corresponding to each pixel value range from the optimization correspondence table; and The pixel value of each pixel coordinate in the optimized image is set to the corresponding optimized pixel value.

Images