Intelligent temperature control system of thermal infrared detection system

Abstract

The invention relates to the technical field of intelligent control, in particular to an intelligent temperature control system of a thermal infrared detection system. The image acquisition unit is used for capturing a temperature distribution image in a target area by using a thermal infrared camera; the image processing unit is used for carrying out preprocessing and windowing processing on the received image and extracting temperature information of the heating target according to a windowing processing result; and the temperature control unit is used for adjusting the temperature of the plurality of heating targets through a method of combining temperature threshold control and multi-target cooperative control based on the temperature information obtained from the image processing unit. According to the system, the position of each heating target can be accurately positioned, the temperature information of each target area can be extracted, the processing efficiency and accuracy are improved, the temperature threshold value of each target is considered, the mutual influence among multiple targets is also considered, and therefore the overall temperature distribution is optimized.

Description

Technical Field

[0001] The invention relates to the field of intelligent control technology, in particular to an intelligent temperature control system of a thermal infrared detection system. Background Art

[0002] The intelligent temperature control system of the thermal infrared detection system operates on a rigid connection platform consisting of a front-end and a back-end platform, where a thermal infrared detection system is installed on the front-end platform and five heating targets are arranged on the back-end platform. The system can use thermal infrared technology to monitor the five heating targets in real time and accurately obtain their coordinate positions in the optical system. Based on these data, the intelligent temperature control system can analyze and adjust the temperature state of each heating target and its surrounding environment to ensure the stable operation of the entire system and achieve the preset temperature control target. It is suitable for application scenarios that require precise temperature management and monitoring, such as industrial automation, safety protection, scientific research and other fields. Although this intelligent temperature control system has shown many advantages, it still faces some challenges in practical applications. In particular, when dealing with closely adjacent or partially overlapping heating targets, in the prior art, the resolution of the thermal infrared camera may not be sufficient to clearly distinguish closely adjacent or partially overlapping targets. If two heating objects are too close, their thermal signals on the thermal image may be mixed together, making it difficult to identify them individually. Some traditional systems use a fixed temperature threshold to distinguish the target from the background. However, in actual application scenarios, due to changes in ambient temperature or temperature fluctuations of the target itself, the fixed threshold may lead to false alarms or missed alarms. Therefore, an intelligent temperature control system for the thermal infrared detection system is designed. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent temperature control system for a thermal infrared detection system to solve the problem raised in the above background technology that if two heating objects are too close, their thermal signals on the thermal image may be mixed together, making it difficult to identify them separately, and some traditional systems use a fixed temperature threshold to distinguish between the target and the background, resulting in false alarms or missed alarms.

[0004] To achieve the above object, the present invention provides an intelligent temperature control system for a thermal infrared detection system, comprising:

[0005] An image acquisition unit, wherein the image acquisition unit is responsible for capturing a temperature distribution image in a target area using a thermal infrared camera, and transmitting the captured image to an image processing unit;

[0006] An image processing unit, the image processing unit is used to perform preprocessing and windowing processing on the received image, and extract the temperature information of the heating target according to the result of the windowing processing; wherein the image processing unit includes an image analysis module and a target recognition module;

[0007] A temperature control unit, the temperature control unit is used to adjust the temperature of several heating targets by combining temperature threshold control with multi-target collaborative control based on the temperature information obtained from the image processing unit. Once an adjustment decision is made, the temperature control unit sends an instruction to the strategy execution unit;

[0008] A policy execution unit is used to execute physical temperature adjustment actions according to instructions issued by the temperature control unit.

[0009] As a further improvement of the technical solution, in the image processing unit:

[0010] The image analysis module is used to analyze and divert the received images, and perform real-time window calculations based on the results fed back by the target recognition module, and update the window position, and calculate the temperature information of the heating target in the updated window area;

[0011] The target recognition module is used to perform non-real-time processing on the received image, identify the fever target, and feed back the position of the identified fever target to the image analysis module.

[0012] As a further improvement of the present technical solution, the image analysis module includes a data diversion module, an interface module and a storage module. The data diversion module can divide the collected image data into three paths. The first path outputs the image data through the interface module; the second path temporarily stores the image data in the storage module and waits for the data request of the target recognition module; the third path transmits the image data to the target recognition module.

[0013] As a further improvement of the technical solution, the specific process of the image processing unit is as follows:

[0014] S1. The target recognition module determines the fuzzy position of the heating target according to the preset initial position, and transmits the fuzzy position of the heating target to the image analysis module. The image analysis module performs window processing according to the fuzzy position of the heating target provided by the target recognition module, calculates the coordinates of the heating target in real time in the windowing area, and extracts the initial temperature information at the same time;

[0015] S2, the target recognition module reads the image data from the storage module, performs non-real-time image processing, uses image segmentation and feature extraction algorithms to identify several fever targets, and calculates the precise coordinates of the several fever targets in the entire image, and then sends the coordinate information of the identified several fever targets back to the image analysis module;

[0016] S3. The image analysis module dynamically adjusts the window position according to the coordinate information of the several heating targets fed back by the target recognition module, and continues to calculate the coordinates and temperature information of the several heating targets in real time in the adjusted window area.

[0017] As a further improvement of the technical solution, the S1 is specifically as follows:

[0018] Assign each pixel within the window a weight based on its temperature:

[0019] ;

[0020] in, For the window Each pixel The weight of is a parameter greater than zero; is the coordinate in the image The temperature value of the pixel at ;

[0021] ;

[0022] ;

[0023] in, The center point of the heating target coordinate; The center point of the heating target coordinate; For Window Pixel coordinates within ; It is the window area of ​​the fuzzy position of the heating target provided by the target recognition module.

[0024] As a further improvement of the technical solution, in S2, the image segmentation and feature extraction algorithm is used to identify a number of heating targets, and the precise coordinates of the several heating targets in the entire image are calculated, as follows:

[0025] The histogram statistical algorithm is used to obtain the distribution of gray values ​​in the image. Figure 2 The foreground and background are distinguished by binarization, and the connected regions are marked using the binarized image to find different target regions. Based on the marked connected regions, the centroid of each connected region is calculated to obtain the location of the heating target.

[0026] As a further improvement of this technical solution, the histogram Figure 2 The values ​​are then quantized to distinguish the foreground from the background, as follows:

[0027] ;

[0028] ;

[0029] ;

[0030] in, is the number of pixels of the foreground; The threshold for distinguishing foreground and background; is the gray level in the histogram frequency of occurrence; is the number of pixels of the background; is the maximum value of gray value; is the total number of pixels in the image; is the total number of pixels in the image;

[0031] , ;

[0032] , ;

[0033] ;

[0034] in, is the grayscale mean of the foreground; is grayscale; is the grayscale mean of the background; is the global grayscale mean of the image;

[0035] Between-class variance Defined as:

[0036] ;

[0037] in, is the between-class variance; is the ratio of foreground pixels to the total number of pixels; is the ratio of background pixels to the total number of pixels;

[0038] Find The largest The value is the optimal threshold :

[0039] ;

[0040] in, To make the between-class variance Maximum threshold; use Binarize the image, that is, divide the pixels in the image into foreground and background.

[0041] As a further improvement of the technical solution, the binarized image is used to mark connected regions to find different target regions, as follows:

[0042] S21. Initialize the marking matrix , all elements are set to 0;

[0043] S22, scanning the binarized image, when encountering an unmarked foreground pixel, assigning a new label number and putting the pixel into a queue;

[0044] S23, taking a pixel from the queue, checking all pixels in the eight neighborhoods of the pixel, and if a neighborhood point is also a foreground pixel and is not marked, marking it and putting it into the queue;

[0045] S24, repeat the steps until the queue is empty, then one connected domain is marked, and continue scanning the image until all connected domains are marked.

[0046] As a further improvement of the technical solution, in the temperature control unit, the temperature of several heating targets is adjusted by combining temperature threshold control with multi-target collaborative control, as follows:

[0047] Dynamic temperature threshold calculation:

[0048] ;

[0049] in, For the goal Dynamic temperature thresholds; is the weight of historical temperature; For the goal The historical temperature average of is the ambient temperature; is the adjustment factor;

[0050] Multi-objective collaborative control:

[0051] ;

[0052] in, To minimize the overall temperature deviation and the temperature difference function between adjacent heating targets; is the number of fever targets; For fever target The current temperature; For the goal The set temperature; is the weight factor; For fever target The set of neighbors of ; For fever target and fever target The temperature influence coefficient between For fever target The current temperature; ;

[0053] ;

[0054] in, is the impact intensity factor; The fever target and fever target The distance between is the impact attenuation factor;

[0055] The control strategy adjusts the temperature of each heat target:

[0056] ;

[0057] in, For fever target The temperature adjustment amount; is the step length;

[0058] The calculated temperature adjustment , adjust the temperature of each target:

[0059] .

[0060] As a further improvement of the present technical solution, the result of the action behavior of the strategy execution unit can be fed back to the temperature control unit in real time, and the temperature control unit can adjust the temperature control according to the result of the action behavior.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1. In the intelligent temperature control system of this thermal infrared detection system, by using the binarized image to mark the connected areas, the intelligent temperature control system can accurately identify and segment different heating targets in the image, which not only helps the system to accurately locate the position of each heating target, but also can extract the temperature information of each target area, improving the processing efficiency and accuracy. In addition, this method helps to count the number of heating targets, realize real-time target tracking, and continuously monitor the position changes of targets in video sequences. By reducing false detections and missed detections, connected area marking enhances the robustness of the system, allowing it to work stably in complex environments;

[0063] 2. In the intelligent temperature control system of the thermal infrared detection system, a more intelligent and adaptive temperature management system can be realized by combining dynamic temperature threshold control and multi-target collaborative control. This combined method not only considers the temperature threshold of each target, but also considers the mutual influence between multiple targets, thereby optimizing the overall temperature distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is the overall flow chart of the present invention;

[0065] Figure 2 is a system block diagram of the image processing unit in the present invention;

[0066] The meaning of each number in the figure is:

[0067] 1. Image acquisition unit; 2. Image processing unit; 21. Image analysis module; 22. Target recognition module; 3. Temperature control unit; 4. Strategy execution unit. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] Example

[0070] See also Figure 1-2 As shown, an intelligent temperature control system of a thermal infrared detection system is provided, comprising an image acquisition unit 1, an image processing unit 2, a temperature control unit 3 and a strategy execution unit 4;

[0071] The image acquisition unit 1 is responsible for capturing the temperature distribution image in the target area using a thermal infrared camera and transmitting the acquired image to the image processing unit 2;

[0072] The image processing unit 2 is used to perform preprocessing and windowing processing on the received image, and extract the temperature information of the heating target according to the result of the windowing processing; wherein the image processing unit 2 includes an image analysis module 21 and a target recognition module 22;

[0073] In image processing unit 2:

[0074] The image analysis module 21 is used to analyze and split the received image, and perform real-time window calculation according to the feedback result of the target recognition module 22, and update the window position, and calculate the temperature information of the heating target in the updated window area;

[0075] The target recognition module 22 is used to perform non-real-time processing on the received image, identify the fever target, and feed back the position of the identified fever target to the image analysis module 21;

[0076] The image analysis module 21 includes a data diversion module, an interface module and a storage module. The data diversion module can divide the collected image data into three paths. The first path outputs the image data through the interface module, and the interface module is a Camera ink interface; the second path temporarily stores the image data in the storage module, and the storage module is a static random access memory SRAM, waiting for the data request of the target recognition module 22; the third path transmits the image data to the target recognition module 22.

[0077] Storing the image data in SRAM instead of processing it immediately can save processing resources of the image analysis module 21, and the high-speed characteristics of SRAM allow data to be stored and read quickly;

[0078] When the target recognition module 22 is ready to process data, it will send a data request to the image analysis module 21. After receiving the request, the image analysis module 21 reads the required data from the SRAM and sends it to the target recognition module 22. In this way, the SRAM acts as a temporary buffer, allowing the image analysis module 21 and the target recognition module 22 to work together more efficiently. The image analysis module 21 can focus on real-time data processing and window calculations, while the target recognition module 22 can focus on complex image processing tasks.

[0079] The specific process of the image processing unit 2 is as follows:

[0080] S1, the target recognition module 22 determines the fuzzy position of the heating target according to the preset initial position, and transmits the fuzzy position of the heating target to the image analysis module 21, the image analysis module 21 performs window processing according to the fuzzy position of the heating target provided by the target recognition module 22, and calculates the coordinates of the heating target in real time in the windowing area, and extracts the initial temperature information at the same time;

[0081] S1 is as follows:

[0082] set up Represents each pixel in the image Given the fuzzy position provided by the target recognition module 22 and a square window around this point Our goal is to find the best point in this window that truly represents the center of the fever target. ;

[0083] Assign each pixel within the window a weight based on its temperature

[0084] ;

[0085] in, For the window Each pixel The weight is based on the temperature value of the pixel calculated; Is a parameter greater than zero, used to control the degree of influence of temperature on weight. , then the weight is directly equal to the temperature value; if , the weight of the high temperature areas will increase significantly, which will help to bias the center point towards these areas; is the coordinate in the image The temperature value of the pixel at , which is a two-dimensional array or matrix, where each element represents the temperature of the corresponding position on the image;

[0086] ;

[0087] ;

[0088] in, The center point of the heating target coordinate; The center point of the heating target coordinate; For Window Pixel coordinates within ; is the fuzzy position of the heating target provided by the target recognition module 22 The window area is usually rectangular or square, and its size can be set according to the actual situation. For example, a square of pixels; the fuzzy position of the heating target provided by the target recognition module 22, which is the starting point for the image analysis module 21 to start searching for the precise position;

[0089] By calculating the weighted centroid in this way, we can get a position that better reflects the center of the high temperature concentration area, rather than simply taking the average value of all pixel coordinates in the window. This can better adapt to the irregular shape of the heating target in actual application scenarios.

[0090] S2, the target recognition module 22 reads the image data from the storage module, performs non-real-time image processing, uses image segmentation and feature extraction algorithms to identify a number of fever targets, and calculates the precise coordinates of the fever targets in the entire image, and then sends the coordinate information of the identified fever targets back to the image analysis module 21;

[0091] In S2, image segmentation and feature extraction algorithms are used to identify several fever targets, and the precise coordinates of the fever targets in the entire image are calculated, as follows:

[0092] The histogram statistical algorithm is used to obtain the distribution of gray values ​​in the image. Figure 2 Binarization is used to distinguish the foreground and background, and the connected regions are marked using the binarized image to find different target regions. Based on the marked connected regions, the centroid of each connected region is calculated to obtain the location of the heating target.

[0093] The grayscale histogram is a function of the grayscale level, describing the number of pixels of this grayscale level in the image (or the frequency of occurrence of pixels of this grayscale level): its horizontal axis is the grayscale level, and the vertical axis represents the number (frequency) of occurrences of this grayscale level in the image.

[0094] The structure of a one-dimensional histogram is expressed as:

[0095]

[0096] in Represents grayscale value For 8-bit grayscale images, the grayscale value The range is 0-255, and the sum of the number of times all pixel values ​​appear is the size of the image.

[0097] The histogram Figure 2 The values ​​are then quantized to distinguish the foreground from the background, as follows:

[0098] ;

[0099] ;

[0100] ;

[0101] in, Foreground (grayscale value less than or equal to )’s number of pixels; The threshold for distinguishing foreground and background; is the gray level in the histogram frequency of occurrence; Background (grey value greater than )’s number of pixels; is the maximum value of the grayscale value, which is 256; is the total number of pixels in the image; is the total number of pixels of the image, and are the width and height of the image respectively;

[0102] , ;

[0103] , ;

[0104] ;

[0105] in, is the grayscale mean of the foreground; is grayscale; is the grayscale mean of the background; is the global grayscale mean of the image;

[0106] Between-class variance Defined as:

[0107] ;

[0108] in, is the inter-class variance, which measures the grayscale difference between foreground and background; is the ratio of foreground pixels to the total number of pixels; is the ratio of background pixels to the total number of pixels; , ; Because variance is a measure of the uniformity of grayscale distribution, the larger the variance, the greater the difference between the two parts of the image. Therefore, traverse from the minimum grayscale value to the maximum grayscale value ,when The between-class variance At maximum, is the optimal segmentation threshold.

[0109] Find The largest The value is the optimal threshold :

[0110] ;

[0111] in, To make the between-class variance The maximum threshold, that is, the optimal threshold; use Binarize the image, that is, divide the pixels in the image into foreground (heat target) and background. Specifically, the grayscale value is less than or equal to Pixels with grayscale values ​​greater than Pixels of are considered as background;

[0112] In the intelligent temperature control system, Figure 2 The value and automatic threshold calculation method can effectively separate the hot target (foreground) from the background in the infrared image. This automatic threshold selection method can adaptively find the best segmentation threshold, thereby improving the accuracy and robustness of target detection;

[0113] The binarized image can be used to extract the temperature information of the heating target. By calculating the average temperature, maximum temperature and temperature standard deviation of the foreground area, more accurate temperature data can be obtained. By analyzing the binarized image, the temperature distribution of the heating target can be understood, so as to carry out more detailed temperature control and monitoring.

[0114] Use the binarized image to mark the connected regions and find different target regions. Through the connected region marking, each independent heating target in the image can be clearly identified, which is very important for the intelligent temperature control system because it needs to accurately know which areas are heating and monitor and control them. The details are as follows:

[0115] S21. Initialize the marking matrix , all elements are set to 0 (indicating unmarked);

[0116] S22, scanning the binarized image, when encountering an unmarked foreground pixel (i.e., a pixel with a gray value of 255), assigning a new tag number (for example, increasing from 1) and placing the pixel into a queue;

[0117] S23, taking a pixel from the queue, checking all pixels in the eight neighborhoods of the pixel (i.e., up, down, left, right, and diagonal directions), if a neighborhood point is also a foreground pixel and is not marked, mark it (using the same mark number) and put it into the queue;

[0118] S24, repeat the steps until the queue is empty, then one connected domain is marked, and continue scanning the image until all connected domains are marked;

[0119] Thus, the labeling matrix Different values ​​in represent different connected regions, each of which corresponds to an independent target;

[0120] By using the binarized image to mark the connected regions, the intelligent temperature control system can accurately identify and segment different heat-generating targets in the image. This process not only helps the system accurately locate the position of each heat-generating target, but also extracts the temperature information of each target area, such as the average temperature, maximum temperature, and temperature standard deviation. Connected region marking simplifies subsequent image processing tasks, such as feature extraction and target classification, and improves processing efficiency and accuracy. In addition, this method helps to count the number of heat-generating targets, achieve real-time target tracking, and continuously monitor the position changes of targets in video sequences. By reducing false detections and missed detections, connected region marking enhances the robustness of the system, enabling it to work stably in complex environments. In short, connected region marking provides a reliable foundation for the intelligent temperature control system, enabling it to perform target detection, temperature monitoring, and real-time response more effectively.

[0121] Based on the results of the connected domain marking, find the coordinates of the centroid of each connected domain area. The centroid of a connected domain The coordinate formula is as follows:

[0122] ;

[0123] ;

[0124] in Indicates The number of pixels in a connected domain, Indicates that the mark is Pixels The column coordinates of Indicates that the mark is Pixels The row coordinates of .

[0125] S3. The image analysis module 21 dynamically adjusts the window position according to the coordinate information of the several heating targets fed back by the target recognition module 22 to ensure that the window area always covers the target, and continues to calculate the coordinates and temperature information of the several heating targets in real time in the adjusted window area.

[0126] The temperature control unit 3 is used to adjust the temperature of several heating targets by combining temperature threshold control with multi-target collaborative control based on the temperature information obtained from the image processing unit 2. Once the adjustment decision is made, the temperature control unit 3 sends an instruction to the strategy execution unit 4;

[0127] Consider the mutual influence between multiple heat generation targets and optimize the overall temperature distribution through coordinated control; for example, if the temperature of one target is too high, it may cause the temperature of adjacent targets to rise, so the temperature of multiple targets needs to be adjusted at the same time;

[0128] In the temperature control unit 3, the temperatures of several heating targets are adjusted by combining temperature threshold control with multi-target collaborative control, as follows:

[0129] Dynamic temperature threshold calculation:

[0130] ;

[0131] in, For the goal Dynamic temperature threshold of is the weight of historical temperature ( ); For the goal The historical temperature averages of is the ambient temperature; The dynamic temperature threshold provides a personalized upper temperature limit for each heating target. This threshold is dynamically adjusted based on the historical temperature of the target, the ambient temperature, and the adjustment factor. This personalized management can better adapt to the characteristics of different targets and environmental changes, and improve the adaptability and robustness of the system. By dynamically adjusting the temperature threshold, the system can handle emergencies and abnormal temperature changes more flexibly, reduce false alarms and unnecessary temperature adjustments; at the same time, in multi-target collaborative control, the dynamic temperature threshold can be used as a constraint to ensure that the temperature of each target does not exceed its personalized threshold. This helps to balance the temperature deviation of a single target and the temperature difference between adjacent targets during the optimization process.

[0132] Multi-objective collaborative control:

[0133] ;

[0134] in, To minimize the overall temperature deviation and the temperature difference function between adjacent heating targets; is the number of fever targets; For fever target The current temperature; For the goal The set temperature; is a weight factor used to balance the temperature deviation of a single target and the temperature difference between adjacent targets; For fever target The set of neighbors of ; For fever target and fever target The temperature influence coefficient between For fever target The current temperature; ;

[0135] The multi-objective coordinated control formulation aims to minimize the overall temperature deviation and the temperature difference between adjacent heating targets. , the system ensures that the current temperature of each heating target is as close to its set temperature as possible, while taking into account the temperature differences between adjacent targets. Weighting factor It is used to balance the temperature deviation of a single target and the temperature difference between adjacent targets, which helps to optimize the overall temperature distribution, reduce local overheating and temperature unevenness, and improve the overall performance and stability of the system.

[0136] ;

[0137] in, is the impact intensity factor; The fever target and fever target The distance between To influence the attenuation factor; this matrix reflects the heat conduction relationship between targets, which helps to consider the temperature influence between adjacent targets in multi-target collaborative control, thereby more accurately optimizing the overall temperature distribution.

[0138] The control strategy adjusts the temperature of each heat target:

[0139] ;

[0140] in, For fever target The temperature adjustment amount; is the step length;

[0141] Control strategy adjustment formula Used to calculate each fever target The first term in the formula is To target Current temperature Push to its set temperature , the second term Used to take into account the temperature differences between adjacent targets. Control the speed of adjustment. With this control strategy, the system is able to gradually adjust the temperature of each target so that it is close to the set temperature and coordinated with other adjacent targets, thereby achieving smoother and more efficient temperature control.

[0142] The calculated temperature adjustment , adjust the temperature of each target:

[0143] ;

[0144] The calculated temperature adjustment , adjust each target Current temperature This simple update formula ensures that the temperature of each target is gradually adjusted according to the control strategy until the ideal temperature distribution is achieved. Through this iterative adjustment, the system can dynamically optimize the temperature of each heat target in real-time operation, improve the system's response speed and control accuracy, and ensure that the temperature of all targets is within a safe and desired range.

[0145] The specific steps are: dynamically adjust the temperature threshold of each target based on historical temperature data and ambient temperature; optimize temperature control by minimizing the overall temperature deviation and the temperature difference between adjacent targets; define a temperature influence matrix to describe the mutual influence between targets; use an optimization algorithm to solve the minimization problem and obtain the temperature adjustment amount for each target; adjust the temperature of each target based on the calculated temperature adjustment amount. By combining dynamic temperature threshold control and multi-target collaborative control, a smarter and more adaptive temperature management system can be achieved. This combined method not only considers the temperature threshold of each target, but also the mutual influence between multiple targets, thereby optimizing the overall temperature distribution.

[0146] The strategy execution unit 4 is used to execute physical temperature adjustment actions according to the instructions issued by the temperature control unit 3, such as turning on or off the heating / cooling equipment, adjusting the fan speed, etc.;

[0147] The results of the action behavior of the strategy execution unit 4 can be fed back to the temperature control unit 3 in real time, and the temperature control unit 3 can adjust the temperature control according to the results of the action behavior to ensure that the temperature of all heating targets is always kept within the set safety range. This closed-loop feedback mechanism improves the response speed and control accuracy of the system, ensures the real-time and effectiveness of temperature control, and thus achieves more stable and efficient temperature management.

[0148] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. An intelligent temperature control system for a thermal infrared detection system, characterized in that: include: An image acquisition unit (1), the image acquisition unit (1) being responsible for capturing a temperature distribution image in a target area using a thermal infrared camera, and transmitting the captured image to an image processing unit (2); An image processing unit (2), the image processing unit (2) being used to perform preprocessing and windowing processing on the received image, and extract temperature information of the heating target according to the result of the windowing processing; wherein the image processing unit (2) comprises an image analysis module (21) and a target recognition module (22); A temperature control unit (3), the temperature control unit (3) being used to adjust the temperatures of a plurality of heating targets by combining temperature threshold control with multi-target collaborative control based on the temperature information obtained from the image processing unit (2); once an adjustment decision is made, the temperature control unit (3) issues an instruction to the strategy execution unit (4); A strategy execution unit (4), wherein the strategy execution unit (4) is used to execute physical temperature adjustment actions according to instructions issued by the temperature control unit (3).

2. The intelligent temperature control system of a thermal infrared detection system according to claim 1, characterized in that: In the image processing unit (2): The image analysis module (21) is used to analyze and split the received image, and perform real-time window calculation according to the feedback result of the target recognition module (22), and update the window position, and calculate the temperature information of the heating target in the updated window area; The target recognition module (22) is used to perform non-real-time processing on the received image, identify the fever target, and feed back the position of the identified fever target to the image analysis module (21).

3. The intelligent temperature control system of a thermal infrared detection system according to claim 2, characterized in that: The image analysis module (21) comprises a data distribution module, an interface module and a storage module. The data distribution module is capable of dividing the collected image data into three paths. The first path outputs the image data through the interface module; the second path temporarily stores the image data in the storage module to wait for a data request from the target recognition module (22); and the third path transmits the image data to the target recognition module (22).

4. The intelligent temperature control system of a thermal infrared detection system according to claim 3, characterized in that: The specific process of the image processing unit (2) is as follows: S1, the target recognition module (22) determines the fuzzy position of the heating target according to a preset initial position, and transmits the fuzzy position of the heating target to the image analysis module (21), the image analysis module (21) performs window processing according to the fuzzy position of the heating target provided by the target recognition module (22), calculates the coordinates of the heating target in real time within the window opening area, and extracts the initial temperature information at the same time; S2, the target recognition module (22) reads the image data from the storage module, performs non-real-time image processing, uses image segmentation and feature extraction algorithms to recognize a number of fever targets, and calculates the precise coordinates of the number of fever targets in the entire image, and then sends the coordinate information of the recognized number of fever targets back to the image analysis module (21); S3, the image analysis module (21) dynamically adjusts the window position according to the coordinate information of the plurality of heating targets fed back by the target recognition module (22), and continues to calculate the coordinates and temperature information of the plurality of heating targets in real time in the adjusted window area.

5. The intelligent temperature control system of a thermal infrared detection system according to claim 4, characterized in that: The S1 is specifically as follows: Assign each pixel within the window a weight based on its temperature: ； in, For the window Each pixel The weight of is a parameter greater than zero; is the coordinate in the image The temperature value of the pixel at ; ； ； in, The center point of the heating target coordinate; The center point of the heating target coordinate; For Window Pixel coordinates within ; It is the window area of ​​the blurred position of the heating target provided by the target recognition module (22).

6. The intelligent temperature control system of a thermal infrared detection system according to claim 5, characterized in that: In S2, a plurality of heating targets are identified using image segmentation and feature extraction algorithms, and the precise coordinates of the plurality of heating targets in the entire image are calculated, as follows: The histogram statistical algorithm is used to obtain the distribution of grayscale values ​​in the image, and the histogram is binarized to distinguish the foreground and background. The binarized image is used to mark the connected areas and find different target areas. Based on the marked connected areas, the centroid of each connected area is calculated to obtain the location of the fever target.

7. The intelligent temperature control system of a thermal infrared detection system according to claim 6, characterized in that: The histogram is binarized to distinguish the foreground and the background, as follows: ； ； ； in, is the number of pixels of the foreground; The threshold for distinguishing foreground and background; is the gray level in the histogram frequency of occurrence; is the number of pixels of the background; is the maximum value of gray value; is the total number of pixels in the image; is the total number of pixels in the image; ， ； ， ； ； in, is the grayscale mean of the foreground; is grayscale; is the grayscale mean of the background; is the global grayscale mean of the image; Between-class variance Defined as: ； in, is the between-class variance; is the ratio of foreground pixels to the total number of pixels; is the ratio of background pixels to the total number of pixels; Find The largest The value is the optimal threshold : ； in, To make the between-class variance Maximum threshold; use Binarize the image, that is, divide the pixels in the image into foreground and background.

8. The intelligent temperature control system of a thermal infrared detection system according to claim 7, characterized in that: The binarized image is used to mark connected regions and find different target regions, as follows: S21. Initialize the marking matrix , all elements are set to 0; S22, scanning the binarized image, when encountering an unmarked foreground pixel, assigning a new label number and putting the pixel into a queue; S23, taking a pixel from the queue, checking all pixels in the eight neighborhoods of the pixel, and if a neighborhood point is also a foreground pixel and is not marked, marking it and putting it into the queue; S24, repeat the steps until the queue is empty, then one connected domain is marked, and continue scanning the image until all connected domains are marked.

9. The intelligent temperature control system of a thermal infrared detection system according to claim 8, characterized in that: In the temperature control unit (3), the temperatures of several heating targets are adjusted by combining temperature threshold control with multi-target collaborative control, specifically as follows: Dynamic temperature threshold calculation: ； in, For the goal Dynamic temperature threshold of is the weight of historical temperature; For the goal The historical temperature average of is the ambient temperature; is the adjustment factor; Multi-objective collaborative control: ； in, To minimize the overall temperature deviation and the temperature difference function between adjacent heating targets; is the number of fever targets; For fever target The current temperature; For the goal The set temperature; is the weight factor; For fever target The set of neighbors of ; For fever target and fever target The temperature influence coefficient between For fever target The current temperature; ; ； in, is the impact intensity factor; The fever target and fever target The distance between is the impact attenuation factor; The control strategy adjusts the temperature of each heat target: ; in, For fever target The temperature adjustment amount; is the step length; The calculated temperature adjustment , adjust the temperature of each target: 。 10. The intelligent temperature control system of a thermal infrared detection system according to claim 9, characterized in that: The result of the action behavior of the strategy execution unit (4) can be fed back to the temperature control unit (3) in real time, and the temperature control unit (3) can adjust the temperature control according to the result of the action behavior.

Images