Fire-fighting method and system applied to medical image machine room

Abstract

The invention provides a fire fighting method and system applied to a medical image machine room, and the method comprises the steps: obtaining three continuous frames of monitoring images, and obtaining a first difference image and a second difference image of adjacent frames according to the three continuous frames of monitoring images; obtaining a target difference image; obtaining an infrared image and a visible light image of the target machine room, and obtaining a fused image according to the infrared image and the visible light image; obtaining coordinate information of all pixel points contained in the motion area according to the target differential image, and separating a target image corresponding to the motion area from the fused image according to the coordinate information of the pixel points; and inputting the target image into the trained fire detection model. According to the invention, a fire disaster can be identified timely and accurately, so that related personnel can take fire-fighting measures timely, and expensive equipment in a medical image machine room is protected from being damaged or loss is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of fire detection, and in particular to a fire protection method and system applied to a medical imaging room. Background Art

[0002] Medical imaging equipment plays a vital role in hospitals. They are used for imaging during diagnosis and treatment, including X-ray machines, CT scanners, MRI, etc. In reality, medical imaging equipment usually uses electricity supply, which may cause fire due to aging wires, short circuits, etc. In addition, some equipment such as X-ray machines will generate heat during use, which is a potential fire hazard.

[0003] However, traditional smoke detectors have a high false alarm rate, and often give false alarms due to circuit anomalies. Manual video surveillance as an auxiliary means has low reliability, large smoke identification limitations, and significant interference from subjective factors, resulting in low accuracy of existing fire detection methods. Summary of the invention

[0004] Based on this, the purpose of the present invention is to propose a fire fighting method and system applied to a medical imaging room, aiming to solve the problem of low detection accuracy of the traditional fire detection method in the room where medical imaging equipment is placed.

[0005] In a first aspect, an embodiment of the present invention provides a fire fighting method applied to a medical imaging room, the method comprising:

[0006] Acquire a monitoring video of the target computer room at a first preset time interval, and extract the monitoring video frame by frame to obtain three consecutive monitoring frames, and acquire a first differential image and a second differential image of adjacent frames according to the three consecutive monitoring frames;

[0007] Acquire a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area;

[0008] Acquire an infrared image and a visible light image of the target computer room, and obtain a fused image according to the infrared image and the visible light image;

[0009] Acquire coordinate information of all pixels included in the motion area according to the target differential image, and separate the target image corresponding to the motion area from the fused image according to the coordinate information of the pixels;

[0010] The target image is input into the trained fire detection model, so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room.

[0011] In some embodiments, the step of acquiring a monitoring video of the target computer room at a first preset time interval, extracting the monitoring video frame by frame to acquire three consecutive monitoring images, and acquiring a first differential image and a second differential image of adjacent frames according to the three consecutive monitoring images comprises:

[0012] Obtaining a first pixel value and a second pixel value of any pixel point with the same coordinates in any two adjacent monitoring images of three consecutive frames, and obtaining a pixel difference value according to the first pixel value and the second pixel value;

[0013] Determining whether the pixel difference value is greater than or equal to a first preset difference threshold;

[0014] If the pixel difference value is greater than or equal to a first preset difference threshold, defining the pixel of the corresponding pixel point as the first preset pixel value;

[0015] If the pixel difference value is less than the first preset difference threshold, the pixel of the corresponding pixel point is defined as a second preset pixel value;

[0016] Until all pixel difference values ​​of the same pixels in any two adjacent frames of images are traversed, a first differential image and a second differential image are obtained.

[0017] In some embodiments, the step of obtaining a first pixel value and a second pixel value of any pixel point with the same coordinates in any two adjacent monitoring images of three consecutive frames, and obtaining a pixel difference value according to the first pixel value and the second pixel value includes:

[0018] The pixel difference value is obtained according to the following formula:

[0019] ΔS t，t-1 (x i ,y i )=S t (x i ,y i )-S t-1 (x i ,y i )

[0020] ΔS t+1，t (x i ,y i )=S t+1 (x i ,y i )-S t (x i ,y i )

[0021] Among them, ΔS t，t-1 (x i ,y i) represents the pixel difference value between the monitoring images of the t-1th frame and the tth frame at the same pixel point i, (x i ,y i ) represents the coordinates of the i-th pixel, S t-1 (x i ,y i ), S t (x i ,y i ), S t+1 (x i ,y i ) represent the pixel value of the i-th pixel in the monitoring image of the t-1th frame, tth frame, and t+1th frame respectively.

[0022] In some embodiments, the step of acquiring a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area, comprises:

[0023] Traversing the pixel values ​​of any identical pixel point in the first differential image and the second differential image, if the pixel values ​​of the identical pixel point in the first differential image and the second differential image are identical, defining the pixel value of the corresponding pixel point as a third preset pixel value;

[0024] If the pixel values ​​of the same pixel point in the first differential image and the second differential image are different, the pixel value of the corresponding pixel point is defined as a fourth preset pixel value;

[0025] The motion area includes all pixel points corresponding to the third preset pixel value, and the background area includes all pixel points corresponding to the fourth preset pixel value.

[0026] In some embodiments, the step of acquiring a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area, comprises:

[0027] The areas of the motion area and the background area are obtained respectively, and the area ratio of the motion area is obtained according to the areas of the motion area and the background area;

[0028] Determining that the area ratio of the motion region is greater than a first preset area threshold;

[0029] If the area ratio of the motion region is greater than the first preset area threshold, an infrared image and a visible light image are extracted according to the monitoring video.

[0030] In some embodiments, the step of acquiring the infrared image and the visible light image of the target computer room and obtaining a fused image according to the infrared image and the visible light image comprises:

[0031] The fused image is obtained according to the following formula:

[0032] R(x i ,y i )=α×V(x i ,y i )+β×I(x i ,y i )

[0033] R(x i ,y i ) indicates that the fused image is at coordinate (x i ,y i ) pixel information, V(x i ,y i ) indicates that the infrared image is at coordinate (x i ,y i ) pixel information, I(x i ,y i ) indicates that the visible light image is at coordinate (x i ,y i ), α represents the target fusion weight of the infrared image, and β represents the target fusion weight of the visible light image.

[0034] In some embodiments, the step of obtaining the fusion weight of the infrared image and the fusion weight of the visible light image includes:

[0035] Set the fusion weight set (α i , β i )=[(0.1,0.9), (0.2,0.8), ..., (0.9,0.8)], obtaining a fusion image obtained under each fusion weight, and obtaining evaluation indicators under each set of fusion weights according to the fusion image obtained under each set of fusion weights, wherein the evaluation indicators include gradient value, mutual information, and standard deviation;

[0036] Sort the evaluation indicators of each group of fusion weights, and obtain the score of each evaluation indicator from the preset scoring table according to the sorting result, and obtain the total score corresponding to each group of fusion weights according to the score of each evaluation indicator:

[0037] F(α i , β i )=a1F G +a2F I +a3F S

[0038] Among them, F(α i , β i ) represents the total score of the fusion weight of the i-th group, a1, a2, a3 represent the weights of the gradient value, mutual information, and standard deviation, respectively, and F G represents the score of the gradient value corresponding to the i-th group of fusion weights, FI represents the score of the mutual information corresponding to the i-th group of fusion weights, F S represents the score of the standard deviation corresponding to the i-th group of fusion weights;

[0039] The maximum total score is screened out, and a set of fusion weights corresponding to the maximum total score is selected as the target fusion weights.

[0040] In some embodiments, the gradient value is obtained according to the following formula:

[0041]

[0042] Among them, G represents the gradient value of the fused image, the size of the fused image is K×L, and f x represents the gradient of the fused image in the x direction, f y Represents the gradient of the fused image in the y direction;

[0043] The mutual information is obtained according to the following formula:

[0044] I=I1+I2

[0045] Where I represents the mutual information of the fused image, I1 represents the mutual information between the fused image and the infrared image, and I2 represents the mutual information between the fused image and the visible light image.

[0046] The standard deviation is obtained according to the following formula:

[0047]

[0048] Where S represents the standard deviation of the fused image, μ represents the average gray value of the fused image, and F(x, y) represents the fused image.

[0049] In some embodiments, the step of inputting the target image into a trained fire detection model so that the fire detection model analyzes the color features, temperature features, and texture features of the motion area to obtain a detection result of the target computer room includes:

[0050] Obtain multiple historical images of fires and non-fires, and mark the fire areas of the historical images of fires, and separate the fire areas according to the marking results to obtain positive samples;

[0051] Historical images in which no fire occurred are defined as negative samples, and the positive samples and negative samples are input into the initial fire detection model for training to obtain a final fire detection model.

[0052] In a second aspect, an embodiment of the present invention further provides a fire protection system applied to a medical imaging room, the system comprising:

[0053] An image extraction module is used to obtain a monitoring video of a target computer room at a first preset time interval, and extract the monitoring video frame by frame to obtain three consecutive frames of monitoring images, and obtain a first differential image and a second differential image of adjacent frames based on the three consecutive frames of monitoring images;

[0054] A differential image acquisition module, configured to acquire a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area;

[0055] An image fusion module, used to acquire an infrared image and a visible light image of the target computer room, and obtain a fused image according to the infrared image and the visible light image;

[0056] A motion region separation module, used for acquiring coordinate information of all pixels included in the motion region according to the target differential image, and separating a target image corresponding to the motion region from the fused image according to the coordinate information of the pixels;

[0057] The detection execution module is used to input the target image into the trained fire detection model so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room.

[0058] In a third aspect, an embodiment of the present invention further provides a storage medium, including one or more programs stored in the storage medium, which, when executed, implements a fire protection method applied to a medical imaging room as described above.

[0059] In a fourth aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising a memory and a processor, wherein:

[0060] The memory is used to store computer programs;

[0061] When the processor is used to execute the computer program stored in the memory, the fire fighting method applied to the medical imaging room as described above is implemented.

[0062] Compared with the prior art, this application has the following advantages:

[0063] 1. This application collects three consecutive frames of images to detect moving targets. If the area ratio of the moving area is greater than the first preset area threshold, it means that there is a moving target. That is, its purpose is to use the motion detection algorithm to identify dynamic elements in the video, such as smoke and flames, so as to effectively eliminate the interference factors of static elements and reduce the false recognition rate.

[0064] 2. Since infrared images are thermal radiation images of objects, they are more prominent in the performance of thermal targets, but the contrast and texture features are insufficient. Visible light images are mainly imaged by reflecting visible light, and their texture details and contrast are more prominent, but the imaging effect of visible light images is poor under conditions such as smoke and night. By fusing the visible light image and infrared image of the room, the fused image has the depth features of the two images, which is conducive to the comprehensive evaluation of the model to evaluate the multiple features of the moving target in the fused image, thereby further improving the detection accuracy.

[0065] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description or will be learned through embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A flowchart of a fire fighting method applied to a medical imaging room proposed by the first embodiment of the present invention;

[0067] Figure 2 It is a structural schematic diagram of a fire protection system applied to a medical imaging room in the second embodiment of the present invention.

[0068] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0069] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0070] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0071] See also Figure 1 , which is a flow chart of a fire fighting method applied to a medical imaging room in a first embodiment of the present invention, the method comprises steps S101 to S105, wherein:

[0072] Step S101: acquiring a monitoring video of a target computer room at a first preset time interval, and extracting the monitoring video frame by frame to obtain three consecutive monitoring frames, and acquiring a first differential image and a second differential image of adjacent frames according to the three consecutive monitoring frames;

[0073] It should be pointed out that by utilizing the movement characteristics of smoke and flames in the early stage of a fire and analyzing three consecutive frames of images, it is possible to accurately lock on the moving targets that may exist in the image, so as to subsequently perform accurate fire detection on the moving targets.

[0074] Specifically, in some embodiments, firstly, a first pixel value and a second pixel value of a pixel point at the same coordinates in any two adjacent monitoring images of three consecutive frames are obtained, and a pixel difference value is obtained according to the first pixel value and the second pixel value;

[0075] Determining whether the pixel difference value is greater than or equal to a first preset difference threshold;

[0076] If the pixel difference value is greater than or equal to a first preset difference threshold, defining the pixel of the corresponding pixel point as the first preset pixel value;

[0077] If the pixel difference value is less than the first preset difference threshold, the pixel of the corresponding pixel point is defined as a second preset pixel value;

[0078] Until all pixel difference values ​​of the same pixels in any two adjacent frames of images are traversed, a first differential image and a second differential image are obtained.

[0079] It should be noted that, in this embodiment, the first preset pixel value and the third preset pixel value are equal to each other, both are 255, and the second preset pixel value and the fourth preset pixel value are equal to each other, both are zero.

[0080] More specifically, the pixel difference value is obtained according to the following formula:

[0081] ΔS t，t-1 (x i ,y i )=S t (x i ,y i )-S t-1 (x i ,y i )

[0082] ΔS t+1，t (x i ,y i )=S t+1 (x i ,y i )-S t (x i ,yi )

[0083] Among them, ΔS t，t-1 (x i ,y i ) represents the pixel difference value between the monitoring images of the t-1th frame and the tth frame at the same pixel point i, (x i ,y i ) represents the coordinates of the i-th pixel, S t-1 (x i ,y i ), S t (x i ,y i ), S t+1 (x i ,y i ) represent the pixel value of the i-th pixel in the monitoring image of the t-1th frame, tth frame, and t+1th frame respectively. The above algorithm can accurately obtain the difference of each pixel in adjacent images.

[0084] In addition, the purpose of setting the first preset time is to be able to continuously analyze the monitoring video in real time so as to respond quickly and in real time.

[0085] Step S102: acquiring a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area;

[0086] It should be noted that, in this step, it is also necessary to traverse the pixel value of any identical pixel point in the first differential image and the second differential image. If the pixel value of the identical pixel point in the first differential image and the second differential image is the same, the pixel value of the corresponding pixel point is defined as the third preset pixel value;

[0087] If the pixel values ​​of the same pixel point in the first differential image and the second differential image are different, the pixel value of the corresponding pixel point is defined as a fourth preset pixel value;

[0088] The motion area includes all pixel points corresponding to the third preset pixel value, and the background area includes all pixel points corresponding to the fourth preset pixel value.

[0089] That is to say, if the pixel value of the same pixel in the first differential image and the second differential image is the same, it means that the information of the pixel is indeed constantly changing. It should be pointed out that this method is different from the traditional method of only analyzing the pixel change value of two adjacent frames of images. Due to the influence of actual factors such as the camera and the computer room environment, if only the pixel change value of two adjacent frames of images is analyzed, the existence of a change means the existence of a target, which will have a very high misjudgment rate, because the stability of the camera's photography, as well as the temperature, humidity, brightness, and air particle size of the computer room will cause the information of the same pixel in two adjacent frames of images to fluctuate.

[0090] In addition, it should be pointed out that after identifying the motion area, in order to further reduce the misjudgment rate, it is necessary to obtain the areas of the motion area and the background area respectively, and obtain the area ratio of the motion area based on the areas of the motion area and the background area;

[0091] Determining that the area ratio of the motion region is greater than a first preset area threshold;

[0092] If the area ratio of the motion area is greater than the first preset area threshold, the infrared image and the visible light image are extracted according to the monitoring video to overcome the situation where only a few pixels have changed, because when a real fire occurs, with the change of smoke and flames, there will be a certain area of ​​the area changing.

[0093] Step S103: acquiring an infrared image and a visible light image of the target computer room, and obtaining a fused image according to the infrared image and the visible light image;

[0094] In this step, the fused image is obtained according to the following formula:

[0095] R(x i ,y i )=α×V(x i ,y i )+β×I(x i ,y i )

[0096] R(x i ,y i ) indicates that the fused image is at coordinate (x i ,y i ) pixel information, V(x i ,y i ) indicates that the infrared image is at coordinate (x i ,y i ) pixel information, I(x i ,y i ) indicates that the visible light image is at coordinate (x i ,y i ), α represents the target fusion weight of the infrared image, and β represents the target fusion weight of the visible light image.

[0097] Since infrared images are thermal radiation images of objects, they are more prominent in the performance of thermal targets, but the contrast and texture features are insufficient. Visible light images are mainly imaged by reflecting visible light, and their texture details and contrast are more prominent, but the imaging effect of visible light images is poor under conditions such as smoke and night. In order to take into account the advantages of both, this embodiment proposes to fuse the two images. In addition, in the actual fusion process, different weights have a greater impact on the fusion effect. In order to further improve the fusion effect and ensure the subsequent detection accuracy, it is also necessary to set a fusion weight set (α i , β i )=[(0.1,0.9),(0.2,0.8),...,(0.9,0.8)], that is, α i +β i =1, and there are 9 groups of pre-set weights in total, and then, a fusion image obtained under each fusion weight is obtained, and an evaluation index under each group of fusion weights is obtained according to the fusion image obtained under each group of fusion weights, wherein the evaluation index includes a gradient value, mutual information, and a standard deviation;

[0098] Sort the evaluation indicators of each group of fusion weights, and obtain the score of each evaluation indicator from the preset scoring table according to the sorting result, and obtain the total score corresponding to each group of fusion weights according to the score of each evaluation indicator:

[0099] F(α i , β i )=a1F G +a2F I +a3F S

[0100] Among them, F(α i , β i ) represents the total score of the fusion weight of the i-th group, a1, a2, a3 represent the weights of the gradient value, mutual information, and standard deviation, respectively, and F G represents the score of the gradient value corresponding to the i-th group of fusion weights, F I represents the score of the mutual information corresponding to the i-th group of fusion weights, F SRepresents the score of the standard deviation corresponding to the i-th group of fusion weights; it should be pointed out that the higher the gradient value, the clearer the image and the more prominent the edge information; the higher the mutual information, the higher the similarity of the two images; the larger the standard deviation, the better the fusion quality of the two images, and since there are 9 groups of fusion weights set, each group corresponds to three evaluation indicators, based on this, the preset score table mainly includes three parts, corresponding to the gradient value, mutual information, and standard deviation, and each preset score table is set with 9 levels, the highest score corresponds to the highest score level, the lowest score corresponds to the lowest score level, and so on, from high to low, the score levels are 100 points, 90 points, 80 points, 70 points, 60 points, 50 points, 40 points, 30 points, and 20 points. In this way, after sorting the score of each indicator, the score of each indicator in each group of fusion weights can be quickly obtained, so as to obtain the total score of each group of fusion weights based on the above formula.

[0101] Finally, the maximum total score is screened out from all the total scores, and a set of fusion weights corresponding to the maximum total score is selected as the target fusion weights, that is, used in the fusion image generation algorithm.

[0102] In some embodiments, the gradient value is obtained according to the following formula:

[0103]

[0104] Among them, G represents the gradient value of the fused image, the size of the fused image is K×L, and f x represents the gradient of the fused image in the x direction, f y Represents the gradient of the fused image in the y direction;

[0105] The mutual information is obtained according to the following formula:

[0106] I=I1+I2

[0107] Where I represents the mutual information of the fused image, I1 represents the mutual information between the fused image and the infrared image, and I2 represents the mutual information between the fused image and the visible light image.

[0108] The standard deviation is obtained according to the following formula:

[0109]

[0110] Where S represents the standard deviation of the fused image, μ represents the average gray value of the fused image, and F(x, y) represents the fused image.

[0111] In summary, by accurately obtaining α and β, the temperature information of the infrared image can be highlighted while retaining the edge and texture details of the visible light image, and the infrared and visible light images can be fused in the best ratio.

[0112] Step S104: acquiring coordinate information of all pixels included in the motion area according to the target differential image, and separating the target image corresponding to the motion area from the fused image according to the coordinate information of the pixels;

[0113] It should be noted that in this step, after identifying the moving area, the coordinate information of all pixel points contained in the moving area can be obtained, so as to separate the area where the moving target is located in turn, so as to speed up the subsequent detection speed, and at the same time enable the detection model to perform targeted identification on it and improve the recognition accuracy.

[0114] In addition, if the area ratio of the motion area is less than or equal to the first preset area threshold, it is considered that the changed pixel points contained in the motion area are too small to indicate the possibility of the existence of targets such as smoke and flames. At this time, the subsequent steps will not be performed, but the subsequent three consecutive frames of images will be collected in a loop for analysis.

[0115] Step S105: inputting the target image into the trained fire detection model, so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room.

[0116] It should be pointed out that in order to obtain the trained fire detection model, it is necessary to obtain multiple historical images of fires and non-fires, and to annotate the fire areas of the historical images of fires. The fire areas are separated according to the annotated results to obtain positive samples.

[0117] Historical images in which no fire occurred are defined as negative samples, and the positive samples and negative samples are input into the initial fire detection model for training to obtain a final fire detection model, which is then used to extract and identify multiple features of the separated motion area, thereby accurately obtaining the final fire identification result.

[0118] In addition, through the above-mentioned precise fire analysis method, it can be ensured that the fire is identified in real time and accurately, and the fire situation can be fed back to the relevant personnel in a timely manner, so that effective fire fighting measures can be taken to protect the expensive equipment in the medical imaging room from or reduce losses.

[0119] In summary, according to the above-mentioned fire-fighting method applied to a medical imaging room, by collecting three consecutive frames of images to detect moving targets, if the area ratio of the moving area is greater than the first preset area threshold, it means that there is a moving target, that is, its purpose is to use a motion detection algorithm to identify dynamic elements in the video, such as smoke and flames, so as to effectively eliminate the interference factors of those static elements and reduce the false recognition rate; in addition, by fusing the visible light image and infrared image of the room, the fused image has the depth features of the two images, which is conducive to the comprehensive and comprehensive evaluation of the model to evaluate multiple features of the moving target in the fused image, thereby further improving the detection accuracy.

[0120] See also Figure 2 , which is a schematic diagram of the structure of a fire protection system applied to a medical imaging room in a second embodiment, the system comprises:

[0121] An image extraction module 10 is used to obtain a monitoring video of a target computer room at a first preset time interval, and extract the monitoring video frame by frame to obtain three consecutive monitoring images, and obtain a first differential image and a second differential image of adjacent frames based on the three consecutive monitoring images;

[0122] A differential image acquisition module 20, configured to acquire a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area;

[0123] An image fusion module 30, used to obtain an infrared image and a visible light image of the target computer room, and obtain a fused image according to the infrared image and the visible light image;

[0124] A motion region separation module 40 is used to obtain coordinate information of all pixels included in the motion region according to the target differential image, and separate a target image corresponding to the motion region from the fused image according to the coordinate information of the pixels;

[0125] The detection execution module 50 is used to input the target image into the trained fire detection model, so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room.

[0126] Another aspect of the present invention further provides a computer storage medium on which one or more programs are stored, and when the programs are executed by a processor, the fire fighting method applied to a medical imaging room is implemented.

[0127] Another aspect of the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to implement the above-mentioned fire fighting method applied to a medical imaging room.

[0128] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0129] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0130] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0131] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0132] The above-described embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A fire fighting method applied to a medical imaging room, characterized in that: The method comprises: Acquire a monitoring video of the target computer room at a first preset time interval, and extract the monitoring video frame by frame to obtain three consecutive monitoring frames, and acquire a first differential image and a second differential image of adjacent frames according to the three consecutive monitoring frames; Acquire a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area; Acquire an infrared image and a visible light image of the target computer room, and obtain a fused image according to the infrared image and the visible light image; Acquire coordinate information of all pixels included in the motion area according to the target differential image, and separate the target image corresponding to the motion area from the fused image according to the coordinate information of the pixels; The target image is input into the trained fire detection model, so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room.

2. The fire fighting method applied to a medical imaging room according to claim 1, characterized in that: The step of acquiring the monitoring video of the target computer room at a first preset time interval, extracting the monitoring video frame by frame to obtain three consecutive monitoring images, and acquiring the first differential image and the second differential image of adjacent frames according to the three consecutive monitoring images comprises: Obtaining a first pixel value and a second pixel value of any pixel point with the same coordinates in any two adjacent monitoring images of three consecutive frames, and obtaining a pixel difference value according to the first pixel value and the second pixel value; Determining whether the pixel difference value is greater than or equal to a first preset difference threshold; If the pixel difference value is greater than or equal to a first preset difference threshold, defining the pixel of the corresponding pixel point as the first preset pixel value; If the pixel difference value is less than the first preset difference threshold, the pixel of the corresponding pixel point is defined as a second preset pixel value; Until all pixel difference values ​​of the same pixels in any two adjacent frames of images are traversed, a first differential image and a second differential image are obtained.

3. The fire fighting method applied to a medical imaging room according to claim 2 is characterized in that: The step of obtaining a first pixel value and a second pixel value of any pixel point with the same coordinates in any two adjacent monitoring images of three consecutive frames, and obtaining a pixel difference value according to the first pixel value and the second pixel value comprises: The pixel difference value is obtained according to the following formula: ΔS t，t-1 (x i ,and i )=S t (x i ,and i )-S t-1 (x i ,and i ) ΔS t+1，t (x i ,and i )=S t+1 (x i ,and i )-S t (x i ,and i ) Among them, ΔS t，t-1 (x i ,y i ) represents the pixel difference value between the monitoring images of the t-1th frame and the tth frame at the same pixel point i, (x i ,y i ) represents the coordinates of the i-th pixel, S t-1 (x i ,y i ), S t (x i ,y i ), S t+1 (x i ,y i ) represent the pixel value of the i-th pixel in the monitoring image of the t-1th frame, tth frame, and t+1th frame respectively.

4. The fire fighting method applied to a medical imaging room according to claim 3 is characterized in that: The step of acquiring a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area, comprises: Traversing the pixel values ​​of any identical pixel point in the first differential image and the second differential image, if the pixel values ​​of the identical pixel point in the first differential image and the second differential image are identical, defining the pixel value of the corresponding pixel point as a third preset pixel value; If the pixel values ​​of the same pixel point in the first differential image and the second differential image are different, the pixel value of the corresponding pixel point is defined as a fourth preset pixel value; The motion area includes all pixel points corresponding to the third preset pixel value, and the background area includes all pixel points corresponding to the fourth preset pixel value.

5. The fire fighting method applied to a medical imaging room according to claim 4 is characterized in that: The step of acquiring a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area, comprises: The areas of the motion area and the background area are obtained respectively, and the area ratio of the motion area is obtained according to the areas of the motion area and the background area; Determining that the area ratio of the motion region is greater than a first preset area threshold; If the area ratio of the motion region is greater than the first preset area threshold, an infrared image and a visible light image are extracted according to the monitoring video.

6. The fire fighting method applied to a medical imaging room according to claim 5, characterized in that: The step of acquiring the infrared image and the visible light image of the target computer room and obtaining a fused image according to the infrared image and the visible light image comprises: The fused image is obtained according to the following formula: R(x i ,and i )=α×V(x i ,and i )+β×I(x i ,and i ) R(x i ,y i ) indicates that the fused image is at coordinate (x i ,y i ) pixel information, V(x i ,y i ) indicates that the infrared image is at coordinate (x i ,y i ) pixel information, I(x i ,y i ) indicates that the visible light image is at coordinate (x i ,y i ), α represents the target fusion weight of the infrared image, and β represents the target fusion weight of the visible light image.

7. The fire fighting method applied to a medical imaging room according to claim 6, characterized in that: The steps of obtaining the fusion weight of the infrared image and the fusion weight of the visible light image include: Set the fusion weight set (α i , β i )=[(0.1,0.9), (0.2,0.8), ..., (0.9,0.8)], obtaining a fusion image obtained under each fusion weight, and obtaining evaluation indicators under each set of fusion weights according to the fusion image obtained under each set of fusion weights, wherein the evaluation indicators include gradient value, mutual information, and standard deviation; Sort the evaluation indicators of each group of fusion weights, and obtain the score of each evaluation indicator from the preset scoring table according to the sorting result, and obtain the total score corresponding to each group of fusion weights according to the score of each evaluation indicator: F(a i ,b i )=a1F G +a2F I +a3F S Among them, F(α i , β i ) represents the total score of the fusion weight of the i-th group, a1, a2, a3 represent the weights of the gradient value, mutual information, and standard deviation, respectively, and F G represents the score of the gradient value corresponding to the i-th group of fusion weights, F I represents the score of the mutual information corresponding to the i-th group of fusion weights, F S represents the score of the standard deviation corresponding to the i-th group of fusion weights; The maximum total score is screened out, and a set of fusion weights corresponding to the maximum total score is selected as the target fusion weights.

8. The fire fighting method applied to a medical imaging room according to claim 7 is characterized in that: The gradient value is obtained according to the following formula: Among them, G represents the gradient value of the fused image, the size of the fused image is K×L, and f x represents the gradient of the fused image in the x direction, f y Represents the gradient of the fused image in the y direction; The mutual information is obtained according to the following formula: I=I1+I2 Where I represents the mutual information of the fused image, I1 represents the mutual information between the fused image and the infrared image, and I2 represents the mutual information between the fused image and the visible light image. The standard deviation is obtained according to the following formula: Where S represents the standard deviation of the fused image, μ represents the average gray value of the fused image, and F(x, y) represents the fused image.

9. The fire fighting method applied to a medical imaging room according to claim 8, characterized in that: The step of inputting the target image into the trained fire detection model so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room includes: Obtain multiple historical images of fires and non-fires, and mark the fire areas of the historical images of fires, and separate the fire areas according to the marking results to obtain positive samples; Historical images in which no fire occurred are defined as negative samples, and the positive samples and negative samples are input into the initial fire detection model for training to obtain a final fire detection model.

10. A fire protection system applied to a medical imaging room, characterized in that: The system comprises: An image extraction module is used to obtain a monitoring video of a target computer room at a first preset time interval, and extract the monitoring video frame by frame to obtain three consecutive frames of monitoring images, and obtain a first differential image and a second differential image of adjacent frames based on the three consecutive frames of monitoring images; A differential image acquisition module, configured to acquire a target differential image according to the first differential image and the second differential image, wherein the target differential image includes a motion area and a background area; An image fusion module, used to acquire an infrared image and a visible light image of the target computer room, and obtain a fused image according to the infrared image and the visible light image; A motion region separation module, used for acquiring coordinate information of all pixels included in the motion region according to the target differential image, and separating a target image corresponding to the motion region from the fused image according to the coordinate information of the pixels; The detection execution module is used to input the target image into the trained fire detection model so that the fire detection model analyzes the color features, temperature features and texture features of the motion area to obtain the detection result of the target computer room.

Images