A fire point video real-time monitoring method and system for mountain fire detection

Abstract

The present invention relates to the field of video compression technology, and in particular to a method and system for real-time monitoring of fire point videos for mountain fire detection, the method comprising: obtaining visible light monitoring videos and infrared monitoring videos; determining a first key frame from the visible light monitoring video according to a first fire point threshold, and determining a second key frame from the infrared monitoring video according to a second fire point threshold; determining the initial importance by using the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame, the second fire point threshold, and the first key frame; correcting the initial importance according to the first temperature data corresponding to the first key frame when it was shot, the first smoke concentration, the second temperature data and the second smoke concentration historically monitored in the target area, and obtaining the final importance; determining a quantization parameter according to the final importance; and compressing the first key frame based on the quantization parameter. In this way, the present invention improves the reliability and flexibility of image compression.

Description

Technical Field

[0001] The present invention relates to the technical field of video compression, and in particular to a fire point video real-time monitoring method and system for mountain fire detection. Background Art

[0002] With the development of science and technology, remote sensing technology, drones, artificial intelligence, and big data analysis are used to monitor forest fires in forest systems, ensuring the reliability and comprehensiveness of forest fire monitoring. Wildfire monitoring is of great significance for disaster prevention and mitigation as well as ecological protection. Effective wildfire monitoring can timely detect and assess fire risks, reduce the damage of fires to forest ecosystems, and protect biodiversity. At the same time, timely warning and response can protect people's lives and reduce economic losses.

[0003] In some scenarios, after monitoring a forest fire, the monitoring video needs to be compressed before transmission. At present, when monitoring the forest fire monitoring video of the forest system, a unified compression parameter is often used to compress the forest fire monitoring video. In this way, different key frames have different importance. Key frames with a higher probability of wildfires should retain higher image quality during compression. Using a unified compression parameter cannot guarantee that the key frames in the monitoring video retain higher image quality, and the reliability and flexibility of image compression are low. Summary of the invention

[0004] In order to solve the technical problems of low reliability and flexibility of image compression, the purpose of the present invention is to provide a fire point video real-time monitoring method and system for mountain fire detection. The technical scheme adopted is as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for real-time monitoring of fire point videos for wildfire detection, comprising: acquiring visible light monitoring video and infrared monitoring video of a target area for wildfire detection; determining a first key frame from the visible light monitoring video according to a first fire point threshold, and determining a second key frame from the infrared monitoring video according to a second fire point threshold; determining an initial importance of content contained in the first key frame using an area of ​​a connected domain in the second key frame, a heat value of the connected domain in the second key frame, the second fire point threshold, and the first key frame; correcting the initial importance according to first temperature data corresponding to when the first key frame was shot, a first smoke concentration, and second temperature data and a second smoke concentration historically monitored in the target area, to obtain a final importance of content contained in the first key frame; determining a quantization parameter for the first key frame when performing video compression according to the final importance, the quantization parameter characterizing image quality after video compression; and performing video compression on the first key frame based on the quantization parameter.

[0006] Optionally, determining a first key frame from the visible light monitoring video according to a first fire point threshold, and determining a second key frame from the infrared monitoring video according to a second fire point threshold includes: determining multiple first marker frames from the visible light monitoring video according to different stages of wildfire monitoring, and determining multiple second marker frames from the infrared monitoring video; calculating a first difference between each pixel point in each of the first marker frames and a pixel point at a corresponding position in a previous frame image of the first marker frame, and superimposing the first difference to obtain a first superposition value, calculating a second difference between each pixel point in each of the second marker frames and a pixel point at a corresponding position in a previous frame image of the second marker frame, and superimposing the second difference, Obtain a second superposition value; select a minimum value from each of the first superposition values ​​as the first fire point threshold, and select a minimum value from each of the second superposition values ​​as the second fire point threshold; when the sum of pixel differences between each pixel point of the first current image frame in the visible light monitoring video and each pixel point of the previous image frame of the first current image frame is greater than the first fire point threshold, determine that the first current image frame is the first key frame; when the sum of differences between the heat value of the second current image frame in the infrared monitoring video and the heat value of the previous image frame of the second current image frame is greater than the second fire point threshold, determine that the second current image frame is the second key frame.

[0007] Optionally, using the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame, the second fire point threshold and the first key frame, determining the initial importance of the content contained in the first key frame includes: clustering the pixels in the second key frame to obtain a clustering result; determining the connected domain in the second key frame according to the clustering result; determining the degree of danger of the fire point represented by the connected domain according to the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame and the second fire point threshold; obtaining frame difference for the first key frame in the visible light monitoring video corresponding to the second key frame according to the degree of danger of the fire point represented by the connected domain to obtain a target frame difference area; determining the initial importance of the content contained in the first key frame according to the area of ​​the target frame difference area corresponding to the first key frame in the visible light monitoring video and the degree of danger of the fire point represented by the connected domain in the second key frame corresponding to the first key frame.

[0008] Optionally, determining the degree of danger of the fire point represented by the connected domain based on the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame and the second fire point threshold includes: calculating the third difference in heat value between the current connected domain and each adjacent connected domain in the second key frame, superimposing each of the third differences to obtain a third superimposed value; when the third superimposed value is greater than or equal to the second fire point threshold, determining that the first product of the area of ​​the current connected domain in the second key frame and the third superimposed value is the initial degree of danger of the fire point represented by the current connected domain; when the third superimposed value is less than the second fire point threshold, determining that the initial degree of danger of the fire point represented by the current connected domain in the second key frame is a preset value; and using the rate of change of the area of ​​the current connected domain in the subsequent two second key frames and the rate of change of the heat value of the current connected domain in the subsequent two second key frames, correcting the initial degree of danger to obtain the degree of danger of the fire point represented by the current connected domain.

[0009] Optionally, the initial danger level is corrected using the rate of change of the area of ​​the current connected domain in two subsequent second key frames and the rate of change of the heat value of the current connected domain in two subsequent second key frames to obtain the danger level of the fire point represented by the current connected domain, including: calculating the second product of the initial danger level, the rate of change of the area and the rate of change of the heat value; normalizing the second product to obtain the danger level of the fire point represented by the current connected domain.

[0010] Optionally, determining the initial importance of the content contained in the first key frame based on the area of ​​the frame difference region corresponding to the first key frame in the visible light monitoring video and the degree of danger of the fire point represented by the connected domain in the second key frame corresponding to the first key frame includes: superimposing the degree of danger of the fire point represented by each connected domain in the second key frame to obtain a fourth superposition value; determining the area difference between the frame difference region corresponding to the current first key frame and the frame difference region corresponding to the next first key frame adjacent to the current first key frame; and determining the third product of the fourth superposition value and the area difference as the initial importance of the content contained in the current first key frame.

[0011] Optionally, according to the danger level of the fire point represented by the connected domain, frame difference is obtained for the first key frame in the visible light monitoring video corresponding to the second key frame to obtain a target frame difference area, including: obtaining a frame difference for the first key frame in the visible light monitoring video that is closest in time to the second key frame and for which the danger level of the fire point represented by the connected domain is less than a first threshold, to obtain a frame difference area; determining the possibility of a wildfire in the first key frame of the visible light monitoring video by using the grayscale value of the frame difference area in the first key frame, the angle of the line connecting the center points of the frame difference areas corresponding to adjacent first key frames, and the wind direction angle when the first key frame was shot; and selecting the frame difference area corresponding to the first key frame whose possibility is less than the second threshold as the target frame difference area.

[0012] Optionally, the initial importance is corrected according to the first temperature data, the first smoke concentration, and the second temperature data and the second smoke concentration historically monitored in the target area corresponding to the first key frame when the first key frame was taken, so as to obtain the final importance of the content contained in the first key frame, including: dividing the second temperature data monitored historically into multiple groups of data, and taking the temperature average of each group of data as the temperature reference value; dividing the second smoke concentration monitored historically into multiple groups of data, and taking the smoke concentration average of each group of data as the smoke concentration reference value; using the first temperature data, the first smoke concentration, the temperature reference value and the smoke concentration reference value to correct the initial importance, so as to obtain the final importance of the content contained in the first key frame.

[0013] Optionally, determining the quantization parameter of the first key frame during video compression according to the final importance includes: normalizing the inverse of the final importance to obtain a normalized value; and determining the fourth product of a predetermined value and the normalized value as the quantization parameter.

[0014] In a second aspect, an embodiment of the present invention provides a fire point real-time video monitoring system for wildfire detection, comprising: a processor and a memory; wherein the memory is used to store computer programs that can be run on the processor; the processor is used to execute the program stored in the memory to implement the steps of the fire point real-time video monitoring method for wildfire detection as mentioned in the first aspect.

[0015] The present invention has the following beneficial effects: first, a visible light monitoring video and an infrared monitoring video of a target area for wildfire detection are obtained; then, a first key frame is determined from the visible light monitoring video according to a first fire point threshold, and a second key frame is determined from the infrared monitoring video according to a second fire point threshold; secondly, the initial importance of the content contained in the first key frame is determined using the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame, the second fire point threshold, and the first key frame; then, the initial importance is corrected according to the first temperature data, the first smoke concentration, the second temperature data and the second smoke concentration historically monitored in the target area corresponding to the time when the first key frame was shot, to obtain the final importance of the content contained in the first key frame; and according to the final importance, a quantization parameter of the first key frame during video compression is determined, the quantization parameter characterizing the image quality after video compression; finally, the first key frame is compressed based on the quantization parameter.

[0016] In this way, the embodiment of the present invention can determine the final importance of the content contained in the key frame in the visible light surveillance video by determining the visible light surveillance video and the infrared surveillance video. The quantization parameter corresponding to each first key frame is determined based on its importance, so that different quantization parameters are used for key frames of different importance to perform video compression. In this way, the image quality of key frames with higher importance is guaranteed after compression, and the reliability and flexibility of image compression are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A flow chart of a method for real-time video monitoring of fire points for wildfire detection provided by one embodiment of the present invention;

[0019] Figure 2 A schematic diagram of an image of a first key frame in a visible light monitoring video provided by an embodiment of the present invention;

[0020] Figure 3 A schematic diagram of an image of a second key frame in an infrared surveillance video provided by an embodiment of the present invention;

[0021] Figure 4 A schematic structural diagram of a fire point video real-time monitoring system for wildfire detection provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a fire point video real-time monitoring method and system for wildfire detection proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0023] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0024] The following is a detailed description of a method and system for real-time video monitoring of fire points for wildfire detection provided by the present invention in conjunction with the accompanying drawings.

[0025] Embodiment 1:

[0026] See also Figure 1 , which shows a flow chart of a method for real-time video monitoring of fire points for mountain fire detection provided by an embodiment of the present invention, including:

[0027] S101, obtaining visible light monitoring video and infrared monitoring video of a target area for wildfire detection.

[0028] Specifically, the target area can be an area with lush vegetation in the forest or any area in the forest. When conducting wildfire detection, you can choose a visible light camera equipped with high-definition visible light, which can have night vision and zoom capabilities to obtain visible light monitoring video of the target area. In addition, you can also choose to have a thermal infrared thermal imaging device, such as an infrared camera, which can capture the temperature changes in the target area to detect the fire source and heat distribution.

[0029] Furthermore, before monitoring wildfires, we should first ensure that the equipment is normal. When checking the equipment, we should ensure that the camera, sensor, memory, power supply, etc. are working properly. We should check and clean the lens and sensor to ensure the image quality. Then determine the target area for monitoring. We can select a suitable target area according to the terrain, vegetation type and historical fire data to ensure a wide field of view so that it can be captured in the first place when encountering wildfires. Secondly, install the selected monitoring equipment at the selected location, and take visible light video images and thermal infrared video images of the monitoring area at the same time. Finally, start the equipment for real-time video monitoring. The visible light camera records images under natural light conditions, and the infrared camera monitors temperature changes. The monitoring cycle is set to 30 minutes to obtain timed images. After the equipment performs real-time monitoring, the collected visible light monitoring video and infrared monitoring video are stored in the local storage device or cloud to ensure data security and integrity. Back up data regularly to prevent data loss due to equipment failure or storage damage.

[0030] S102, determining a first key frame from the visible light monitoring video according to a first fire point threshold, and determining a second key frame from the infrared monitoring video according to a second fire point threshold.

[0031] Specifically, the embodiment of the present invention needs to divide the visible light monitoring video and the infrared monitoring video into different frames and find the key frames. Among them, the types of the first key frame and the second key frame are both I frames. For example, Figure 2 and Figure 3 As shown, Figure 2 A schematic diagram of an image of a first key frame in a visible light monitoring video provided by an embodiment of the present invention, Figure 3 A schematic diagram of an image of a second key frame in an infrared surveillance video provided by an embodiment of the present invention. Figure 2 and Figure 3 It can be seen that when a wildfire occurs, the wildfire situation can be reflected in both the first key frame and the second key frame.

[0032] Furthermore, when determining the first key frame and the second key frame, as an optional embodiment of the present invention, first determine multiple first marker frames from the visible light monitoring video and determine multiple second marker frames from the infrared monitoring video according to different stages of wildfire monitoring; then calculate the first difference between each pixel point in each first marker frame and the pixel point at the corresponding position in the previous frame image of the first marker frame, and superimpose the first difference to obtain a first superimposed value, calculate the second difference between each pixel point in each second marker frame and the pixel point at the corresponding position in the previous frame image of the second marker frame, and superimpose the second difference to obtain a second superimposed value. ; Then select the minimum value from each first superposition value as the first fire point threshold, and select the minimum value from each second superposition value as the second fire point threshold; secondly, when the sum of the pixel differences between each pixel point of the first current image frame in the visible light monitoring video and each pixel point of the previous image frame of the first current image frame is greater than the first fire point threshold, determine that the first current image frame is the first key frame; finally, when the sum of the differences between the heat value of the second current image frame in the infrared monitoring video and the heat value of the previous image frame of the second current image frame is greater than the second fire point threshold, determine that the second current image frame is the second key frame.

[0033] Specifically, in the historical videos related to wildfires, various stages of wildfires that can be monitored are found, including but not limited to the initial stage, the development stage, the fierce stage, the decay stage, and the extinguishing stage. There are frames with division marks between the various stages, which are recorded as the first mark frame and the second mark frame in the embodiment of the present invention, and the first mark frame and the second mark frame represent different stages of the wildfire.

[0034] Furthermore, when calculating the first fire point threshold, the embodiment of the present invention may use the following formula:

[0035]

[0036] In the above formula, represents the first fire point threshold, Indicates that in the first marker frame The first difference between a pixel point and a pixel point at a corresponding position in the previous frame image. It indicates that the first difference values ​​corresponding to each pixel point in the first marker frame are superimposed.

[0037] Furthermore, when calculating the second fire point threshold, the embodiment of the present invention may use the following formula:

[0038]

[0039] In the above formula, represents the second fire point threshold, Indicates that in the second marker frame The second difference between a pixel point and a pixel point at a corresponding position in the previous frame image. It indicates that the second difference values ​​corresponding to each pixel point in the second marker frame are superimposed.

[0040] Furthermore, when the sum of the pixel differences between the image frame in the visible light monitoring video and the pixel points of the previous image frame is greater than the first fire point threshold, it indicates that there may be a fire point in the image frame, and it is used as the first key frame. When the sum of the heat value differences between the image frame in the infrared monitoring video and the previous image frame is greater than the second fire point threshold, it indicates that there may be a fire point in the image frame, and it is used as the second key frame.

[0041] S103, determining the initial importance of the content included in the first key frame by using the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame, the second fire point threshold, and the first key frame.

[0042] Specifically, the connected domain in the second key frame can be determined according to the clustering result after clustering the pixels in the second key frame. The heat value of the connected domain in the second key frame can be the average of the heat values ​​of different positions in the connected domain. Wherein, when determining the initial importance of the content contained in the first key frame, as an optional embodiment of the present invention, firstly, the pixels in the second key frame are clustered to obtain the clustering result; secondly, the connected domain in the second key frame is determined according to the clustering result; then, the degree of danger of the fire point represented by the connected domain is determined according to the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame and the second fire point threshold; and according to the degree of danger of the fire point represented by the connected domain, the first key frame in the visible light monitoring video corresponding to the second key frame is obtained by frame difference to obtain the target frame difference area; finally, according to the area of ​​the target frame difference area corresponding to the first key frame in the visible light monitoring video and the degree of danger of the fire point represented by the connected domain in the second key frame corresponding to the first key frame, the initial importance of the content contained in the first key frame is determined.

[0043] Specifically, in the embodiment of the present invention, the k-means clustering algorithm can be used to cluster the pixels in the second key frame, so as to classify the pixels in each second key frame in the infrared surveillance video, thereby further dividing the connected domain. The specific implementation method of dividing the connected domain according to the clustering result can be referred to the prior art, and the embodiment of the present invention will not be repeated here.

[0044] Further, when the difference in heat value between a certain connected domain and the surrounding connected domains is large enough, it is considered that it may belong to a fire point. When the area is larger and the temperature difference with the surrounding connected domains is larger, it is considered that the degree of danger of being a fire point is higher, that is, the possibility of being a fire point is greater. When determining the degree of danger of a connected domain representing a fire point, as an optional embodiment of the present invention, the third difference in heat value between the current connected domain and each adjacent connected domain in the second key frame is calculated, and each third difference is superimposed to obtain a third superposition value; when the third superposition value is greater than or equal to the second fire point threshold, the first product of the area of ​​the current connected domain in the second key frame and the third superposition value is determined as the initial degree of danger of the fire point represented by the current connected domain; when the third superposition value is less than the second fire point threshold, the initial degree of danger of the fire point represented by the current connected domain in the second key frame is determined to be a preset value; the initial degree of danger is corrected by using the rate of change of the area of ​​the current connected domain in the subsequent two second key frames and the rate of change of the heat value of the current connected domain in the subsequent two second key frames to obtain the degree of danger of the fire point represented by the current connected domain.

[0045] Specifically, the preset value can be determined according to actual conditions. In the embodiment of the present invention, the preset value is 0.2. The following example is used to illustrate the degree of danger of the fire point represented by the current connected domain. In the embodiment of the present invention, the following formula is used to calculate the initial degree of danger of the fire point represented by the connected domain:

[0046]

[0047] In the above formula, Indicates the second key frame The connected domains represent the initial danger level of the fire point. Indicates the second key frame The area of ​​a connected domain. Indicates the second key frame The connected domain and the surrounding adjacent The third difference in heat values ​​between connected domains. represents the second fire point threshold, Indicates the second key frame The cumulative value of the third difference of the heat value between a connected domain and all the adjacent connected domains. Adjacent connected domains are connected domains with overlapping edges. Indicates a preset value.

[0048] Furthermore, the calculation method of the rate of change of the area of ​​the current connected domain in the two subsequent second key frames can be: the ratio of the absolute value of the difference in the area of ​​the current connected domain in the two subsequent second key frames to the area of ​​the current connected domain in the first subsequent second key frame. And the calculation method of the rate of change of the heat value of the current connected domain in the two subsequent second key frames can be: the ratio of the absolute value of the difference in the heat value of the current connected domain in the two subsequent second key frames to the heat value of the current connected domain in the first subsequent second key frame.

[0049] Furthermore, after obtaining the initial danger level, the initial danger level is corrected. As an optional embodiment of the present invention, the second product of the initial danger level, the rate of change of the area, and the rate of change of the heat value is first calculated; then the second product is normalized to obtain the current connected domain representing the danger level of the fire point.

[0050] Specifically, when the rate of change of the area of ​​the fire point in the second key frame is faster and the rate of change of the temperature is faster, it can be considered that the degree of danger of the connected domain of the fire point is higher. In the embodiment of the present invention, the following formula can be used to calculate the degree of danger of the fire point represented by the current connected domain:

[0051]

[0052] In the above formula, Indicates the second key frame The connected domains represent the danger level of the fire point. Indicates the second key frame The connected domains represent the initial danger level of the fire point. Indicates The rate of change of the area of ​​a connected domain in the two subsequent second key frames. Indicates The rate of change of the heat value of a connected domain in the next two second key frames. Represents the normalization function, which is used to Perform normalization.

[0053] Furthermore, when determining the initial importance of the content contained in the first key frame, as an optional embodiment of the present invention, first, the danger levels of the fire points represented by each connected domain in the second key frame are superimposed to obtain a fourth superposition value; secondly, the area difference between the frame difference region corresponding to the current first key frame and the frame difference region corresponding to the next first key frame adjacent to the current first key frame is determined; finally, the third product of the fourth superposition value and the area difference is determined as the initial importance of the content contained in the current first key frame.

[0054] Specifically, the embodiment of the present invention is based on the embodiment of the present invention. Taking the first key frame as the current first key frame as an example, the following formula is used to calculate the initial importance of the content contained in the current first key frame:

[0055]

[0056] In the above formula, Indicates the number of visible light surveillance videos. The first keyframe contains the initial importance of the content. Indicates the number of visible light surveillance videos. The area difference between the frame difference region obtained from a first key frame and the frame difference region corresponding to the next adjacent first key frame. Indicates the number of visible light surveillance videos. The first key frame corresponds to the second key frame The connected domains represent the danger level of the fire point. Indicates the number of visible light surveillance videos All connected domains in the second key frame corresponding to the first key frame represent the cumulative value of the danger level of the fire point.

[0057] Furthermore, when determining the target frame difference area, as an optional embodiment of the present invention, first, a frame difference is obtained for the first key frame in the visible light monitoring video that is closest in time distance to the second key frame and whose connected domain indicates that the danger level of the fire point is less than a first threshold, to obtain a frame difference area; then, the possibility of a wildfire existing in the first key frame of the visible light monitoring video is determined by using the grayscale value of the frame difference area in the first key frame, the angle of the line connecting the center points of the frame difference areas corresponding to adjacent first key frames, and the wind direction angle when the first key frame was shot; finally, the frame difference area corresponding to the first key frame whose possibility is less than the second threshold is selected as the target frame difference area.

[0058] Specifically, the first threshold and the second threshold can be taken according to actual conditions. In the embodiment of the present invention, the first threshold is taken as 0.3, and the second threshold is taken as 0.2. In the embodiment of the present invention, the first key frame in the visible light monitoring video, which is closest in time to the second key frame and whose connected domain indicates that the danger level of the fire point is less than 0.3, is obtained by frame difference, and the obtained frame difference area is regarded as a smoke or fire point area, so as to calculate the possibility of wildfire. The whiter the frame difference area is, that is, the higher the gray value is, and the more consistent the moving direction is with the wind direction, the more likely it is to represent a smoke area.

[0059] Furthermore, in the embodiment of the present invention, the possibility that a wildfire exists in the first key frame of the visible light monitoring video can be calculated using the following formula:

[0060]

[0061] In the above formula, Indicates the number of visible light surveillance videos. The first keyframe has the possibility of a wildfire. Indicates the number of visible light surveillance videos. The grayscale value of the frame difference area corresponding to the first key frame in the original image. Indicates the number of visible light surveillance videos. The angle between the center point of the frame difference area obtained by a first key frame and the center point of the frame difference area obtained by the next adjacent first key frame. Indicates the above The wind direction angle when the first key frame was taken.

[0062] S104, correcting the initial importance according to the first temperature data and the first smoke density corresponding to the shooting of the first key frame, and the second temperature data and the second smoke density historically monitored in the target area, to obtain the final importance of the content contained in the first key frame.

[0063] Specifically, the temperature data and smoke data in the embodiment of the present invention can be obtained by the sensor. The first temperature data and the first smoke data are collected by the sensor when shooting the first key frame, and the second temperature data and the second smoke density are historical data monitored by the sensor. When correcting the initial importance, as an optional embodiment of the present invention, the second temperature data monitored historically is first divided into multiple groups of data, and the temperature mean of each group of data is used as a temperature reference value;

[0064] Then the historically monitored second smoke concentration is divided into multiple groups of data, and the average smoke concentration of each group of data is used as the smoke concentration reference value; finally, the initial importance is corrected using the first temperature data, the first smoke concentration, the temperature reference value and the smoke concentration reference value to obtain the final importance of the content contained in the first key frame.

[0065] Specifically, the temperature data of different air temperatures monitored in history are divided into multiple data groups, and the temperature mean value in each data group is used as the temperature reference value of each data group. The different smoke concentrations monitored in history are divided into multiple data groups, and the smoke concentration mean value in each data group is used as the smoke concentration reference value. The data group where the first temperature data is located is determined, and the temperature reference value of the data group is determined to be the reference value of the first temperature data. The data group where the first smoke concentration is located is determined, and the smoke concentration reference value of the data group is determined to be the reference value of the first smoke concentration.

[0066] Furthermore, the embodiment of the present invention uses the following formula to correct the initial importance:

[0067]

[0068] In the above formula, The final result is The final importance of the content contained in the first keyframe. Indicates the number of visible light surveillance videos. The first keyframe contains the initial importance of the content. Indicates The first key frame is captured The reading of a related temperature sensor is the first temperature data. Indicates The first key frame is captured The temperature reference value of the relevant temperature sensor. Indicates The first key frame is captured The rate of increase of the readings of the relevant temperature sensors within the last 10 seconds. Indicates The first key frame is captured The reading of a related smoke sensor is the first smoke density data. Indicates The first key frame is captured The smoke density reference value of each relevant smoke sensor. Indicates The first key frame is captured The rate of increase of the readings of the relevant smoke sensors in the last 10 seconds, Indicates that the relevant data of all related smoke sensors are accumulated.

[0069] S105, determining a quantization parameter of the first key frame when performing video compression according to the final importance, where the quantization parameter represents image quality after the video compression.

[0070] Specifically, in the embodiment of the present invention, the quantization parameter may be a QP value, which is an important parameter affecting image quality, and a lower QP value will retain more image details. When determining the quantization parameter of the first key frame during video compression, as an optional embodiment of the present invention, the inverse of the final importance is first normalized to obtain a normalized value; then the fourth product of the predetermined value and the normalized value is determined as the quantization parameter.

[0071] The predetermined value can be set according to actual conditions. In the embodiment of the present invention, the value is 51. The quantization parameter can be calculated by the following formula in the embodiment of the present invention:

[0072]

[0073] In the above formula, Indicates The QP value corresponding to the first key frame, the value range is 0~51. The final result is The final importance of the content contained in the first keyframe. Represents the normalization function, which is used to Perform normalization processing, Indicates rounding up.

[0074] S106: Perform video compression on the first key frame based on the quantization parameter.

[0075] Specifically, after obtaining the quantization parameter QP value, when the encoder performs video compression on the first key frame, the QP value of the first key frame is dynamically adjusted to the value determined by the above embodiment of the present invention using a power control algorithm. After the compression is completed, the encoded video can be analyzed to evaluate the effects of different QP value settings. It is worth noting that the QP value can also be fine-tuned according to actual conditions to further optimize the quality of the video after compression.

[0076] The embodiment of the present invention can determine the final importance of the content contained in the key frame in the visible light surveillance video by determining the visible light surveillance video and the infrared surveillance video. The quantization parameter corresponding to each first key frame is determined based on its importance, so that different quantization parameters are used for key frames of different importance to perform video compression. In this way, the image quality of the key frames with higher importance after compression is guaranteed, and the reliability and flexibility of image compression are improved.

[0077] Embodiment 2:

[0078] Corresponding to the fire point video real-time monitoring method for mountain fire detection provided in the above embodiment, based on the same technical concept, the embodiment of the present invention also provides a fire point video real-time monitoring system for mountain fire detection, and the fire point video real-time monitoring system for mountain fire detection is used to execute the fire point video real-time monitoring method for mountain fire detection. Figure 4 A schematic diagram of a fire point video real-time monitoring system for wildfire detection provided by an embodiment of the present invention is shown in FIG. Figure 4 The fire point video real-time monitoring system for wildfire detection may have relatively large differences due to different configurations or performances, and may include one or more processors 401 and memory 402, the memory 402 is used to store computer programs that can be run on the processor 401, and the processor 401 is used to execute the program stored in the memory 402 to achieve the above Figure 1The various steps in the method embodiment. Among them, the memory 402 can be a temporary storage or a permanent storage. The application stored in the memory 402 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the real-time video monitoring system of the fire point for wildfire detection.

[0079] Furthermore, the processor 401 can be configured to communicate with the memory 402 to execute a series of computer executable instructions in the memory 402 on the fire point video real-time monitoring system for wildfire detection. The fire point video real-time monitoring system for wildfire detection can also include one or more power supplies 403, one or more wired or wireless network interfaces 404, one or more input and output interfaces 405, and one or more keyboards 406.

[0080] Specifically in this embodiment, the fire point video real-time monitoring system for wildfire detection includes a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other through the bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to achieve the above Figure 1 The various steps in the method embodiment are similar to those in the method embodiment, and have the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of the present invention will not be described again here.

[0081] It should be noted that the fire point real-time video monitoring system for wildfire detection provided in the embodiment of the present invention and the fire point real-time video monitoring method for wildfire detection provided in the embodiment of the present invention are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned fire point real-time video monitoring method for wildfire detection, and has the same or similar beneficial effects, and the repeated parts will not be repeated.

[0082] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A fire point video real-time monitoring method for wildfire detection, characterized in that: Acquire visible light surveillance video and infrared surveillance video of the target area for wildfire detection; Determine a first key frame from the visible light monitoring video according to a first fire point threshold, and determine a second key frame from the infrared monitoring video according to a second fire point threshold; Determine the initial importance of the content contained in the first key frame by using the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame, the second fire point threshold, and the first key frame; The initial importance is corrected according to the first temperature data and the first smoke density corresponding to the shooting of the first key frame, the second temperature data and the second smoke density monitored historically in the target area, to obtain the final importance of the content contained in the first key frame; Determine, according to the final importance, a quantization parameter of the first key frame when performing video compression, wherein the quantization parameter represents image quality after the video compression; Video compression is performed on the first key frame based on the quantization parameter.

2. The method for real-time video monitoring of fire points for mountain fire detection according to claim 1, characterized in that: The determining of the first key frame from the visible light monitoring video according to the first fire point threshold, and the determining of the second key frame from the infrared monitoring video according to the second fire point threshold comprises: Determine a plurality of first marker frames from the visible light monitoring video and a plurality of second marker frames from the infrared monitoring video according to different stages of wildfire monitoring; Calculating a first difference between each pixel point in each of the first marker frames and a pixel point at a corresponding position in an image frame before the first marker frame, and superimposing the first difference values ​​of all the pixels in the first marker frame to obtain a first superimposed value, calculating a second difference between each pixel point in each of the second marker frames and a pixel point at a corresponding position in an image frame before the second marker frame, and superimposing the second difference values ​​of all the pixels in the second marker frame to obtain a second superimposed value; Selecting a minimum value from each of the first superposition values ​​as the first fire point threshold, and selecting a minimum value from each of the second superposition values ​​as the second fire point threshold; When the sum of pixel differences between each pixel point of a first current image frame in the visible light monitoring video and each pixel point of an image frame before the first current image frame is greater than the first fire point threshold, determining that the first current image frame is the first key frame; When the sum of the differences between the heat value of the second current image frame in the infrared surveillance video and the heat value of the previous image frame of the second current image frame is greater than the second fire point threshold, the second current image frame is determined to be the second key frame.

3. The fire point video real-time monitoring method for mountain fire detection according to claim 1 is characterized in that: The determining the initial importance of the content included in the first key frame by using the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame, the second fire point threshold, and the first key frame includes: Clustering the pixels in the second key frame to obtain a clustering result; Determine a connected domain in the second key frame according to the clustering result; Determine the danger level of the fire point represented by the connected domain according to the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame and the second fire point threshold; According to the danger level of the fire point represented by the connected domain, a first key frame in the visible light monitoring video corresponding to the second key frame with the closest time distance is obtained to obtain a target frame difference area; The initial importance of the content contained in the first key frame is determined based on the area of ​​the target frame difference region corresponding to the first key frame in the visible light monitoring video and the connected domain in the second key frame corresponding to the first key frame representing the danger level of the fire point.

4. The method for real-time video monitoring of fire points for mountain fire detection according to claim 3, characterized in that: Determining the danger level of the fire point represented by the connected domain according to the area of ​​the connected domain in the second key frame, the heat value of the connected domain in the second key frame and the second fire point threshold comprises: Calculating a third difference in heat value between the current connected domain and each of the adjacent connected domains in the second key frame, and superimposing each of the third difference values ​​to obtain a third superposition value; In the case where the third superposition value is greater than or equal to the second fire point threshold, determining a first product of an area of ​​a current connected domain in the second key frame and the third superposition value as an initial danger level of a fire point represented by the current connected domain; In the case where the third superposition value is less than the second fire point threshold, determining that the initial danger level of the fire point represented by the current connected domain in the second key frame is a preset value; The initial danger level is corrected by using the rate of change of the area of ​​the current connected domain in the two subsequent second key frames and the rate of change of the heat value of the current connected domain in the two subsequent second key frames to obtain the danger level of the fire point represented by the current connected domain.

5. The method for real-time video monitoring of fire points for mountain fire detection according to claim 4, characterized in that: The method of modifying the initial danger level by using the change rate of the area of ​​the current connected domain in the two subsequent second key frames and the change rate of the heat value of the current connected domain in the two subsequent second key frames to obtain the danger level of the fire point represented by the current connected domain includes: Calculating a second product of the initial danger level, the rate of change of the area, and the rate of change of the heat value; The second product is normalized to obtain the current connected domain representing the danger level of the fire point.

6. The method for real-time video monitoring of fire points for mountain fire detection according to claim 3, characterized in that: The determining the initial importance of the content contained in the first key frame according to the area of ​​the target frame difference region corresponding to the first key frame in the visible light monitoring video and the connected domain in the second key frame corresponding to the first key frame represents the danger level of the fire point, comprises: superimposing the danger levels of the fire points represented by the connected domains in the second key frame to obtain a fourth superimposed value; Determine an area difference between a target frame difference region corresponding to a current first key frame and a target frame difference region corresponding to a subsequent first key frame adjacent to the current first key frame; A third product of the fourth superposition value and the area difference value is determined as the initial importance of the content included in the current first key frame.

7. The method for real-time video monitoring of fire points for wildfire detection according to any one of claims 1 to 6, characterized in that: The initial importance is corrected according to the first temperature data and the first smoke density corresponding to the shooting of the first key frame, and the second temperature data and the second smoke density historically monitored in the target area to obtain the final importance of the content contained in the first key frame, including: Dividing the historically monitored second temperature data into multiple groups of data, and taking the temperature mean of each group of data as a temperature reference value; Dividing the second smoke concentration monitored historically into multiple groups of data, and taking the smoke concentration average of each group of data as a smoke concentration reference value; The initial importance is corrected using the first temperature data, the first smoke density, the temperature reference value, and the smoke density reference value to obtain a final importance of the content included in the first key frame.

8. The method for real-time video monitoring of fire points for wildfire detection according to any one of claims 1 to 6, characterized in that: Determining the quantization parameter of the first key frame when performing video compression according to the final importance includes: Normalizing the reciprocal of the final importance to obtain a normalized value; A fourth product of a predetermined value and the normalized value is determined as the quantization parameter.

9. A fire point video real-time monitoring system for wildfire detection, characterized in that: The fire point video real-time monitoring system for wildfire detection includes: a processor and a memory; wherein the memory is used to store computer programs that can be run on the processor; the processor is used to execute the program stored in the memory to implement the steps of the fire point video real-time monitoring method for wildfire detection as described in any one of claims 1-8.

Images