A method for locating and restoring fire points in surveillance videos

Abstract

The present invention relates to the technical field of fire scene investigation, and specifically to a method for on-site positioning and restoration of a fire point via a surveillance video. The method comprises the following steps: exporting a surveillance video into a video format recognizable by a PC, searching for characteristic frames, recording characteristic frames where flames or smoke first appear, converting the frames into picture A, recording coordinates of the leftmost and rightmost pixels of the flame, connecting the method to an on-site positioning device to collect online video and read the video of a surveillance camera, drawing an on-site position map, determining the closest and farthest possible distances between the flame position at the fire scene and the camera, finding an on-site position corresponding to these coordinate points and marking it in a plane map, and marking the range of the fire point in the video, i.e., the range formed by the projection of all possible positions onto the ground. Compared with the prior art, the present invention calculates and exports the fire coordinate points in the original video, loads the video signal of the video camera at the same position at the scene into the device, locates the fire point at the actual scene, and thus achieves high-precision positioning at the actual scene.

Description

Technical Field

[0001] The invention relates to the technical field of fire scene investigation, in particular to a method for on-site positioning and restoration of a fire point via a monitoring video. Background Art

[0002] Around the world, fire is one of the major disasters faced by mankind, causing a large number of casualties and property losses. With the development of science and technology, fire warning technology and fire fighting technology have made great progress in recent years. However, serious and major fire accidents still occur from time to time. After a fire occurs, it is crucial to accurately locate the fire point and accurately analyze the cause of the fire. On the one hand, accurate location of the fire point can be used as evidence to clarify the responsibility for the fire accident; on the other hand, accurate analysis of the cause of the fire is conducive to early warning and prevention of similar fires in the future.

[0003] Currently, fire video investigations are mainly conducted based on experience. There is no accurate pixel-level video surveillance on-site investigation method, and the fire point is located within the video, resulting in low technical problems in the accuracy of on-site positioning.

[0004] Therefore, it is necessary to design a method for on-site positioning and restoration of the fire point in surveillance video, detect the initial position of the fire in the video, and use the on-site positioning device and on-site camera to calibrate the fire scene position, so as to obtain the precise location of the scene corresponding to the target pixel point. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the existing technology and provide a method for on-site positioning and restoration of the fire point in a monitoring video. The method calculates and derives the fire coordinate point in the original video, and then loads the video signal of the video camera at the same position on the scene into the device to locate the fire point at the actual scene, thereby achieving high-precision positioning on the actual scene.

[0006] In order to achieve the above object, the present invention provides a method for locating and restoring the fire point on-site using a surveillance video, comprising the following steps:

[0007] S1: Copy the video footage captured by the monitoring equipment around the fire scene to a computer;

[0008] S2: Use flame features or smoke features to search for feature frames in all frames of the video, record the feature frames where flames or smoke first appear, and convert them into images;

[0009] S3: When abnormally bright pixels appear continuously in the image in S2, the continuity includes the continuity of time and the continuity of pixel coordinates, and the frame that initially produces the abnormal brightness is found as the initial frame. The coordinates of all pixels on the leftmost and rightmost sides of the flame in the image in S2 are recorded;

[0010] S4: exporting the coordinates of all flames or smoke in the initial frame in S3 and importing them into the on-site positioning device through a computer interface program;

[0011] S5: The coordinates in S4 are superimposed on the corresponding positions of the online video using a special color by a computer, so as to obtain the effect of the coordinates being marked on the video, or the transparency of the image in S2 is modified in real time, so that the modified image is superimposed on the collected online video to form a new online video. The online video observes the special color position in real time, and the camera position is corrected according to the position of an object fixed at the fire scene before and after the fire, so that the modified image and the new online video overlap based on the position of the object;

[0012] S6: Position the camera at the actual scene based on the flame or smoke obstruction position, including the camera position and the corresponding scene position of the coordinate point at a certain elevation, and determine the closest and farthest possible distances between the flame position at the fire scene and the camera;

[0013] S7: Find a location on the actual site corresponding to the coordinate point in S6 and mark it on the plan;

[0014] S8: The maximum range that these points can form in the plan view in S7 is the possible range of the fire point. That is, according to the on-site location map, mark the range of the fire point in the video. The marked range of the fire point is the range formed by the projection of all positions onto the ground.

[0015] The on-site positioning device includes a surveillance camera address resolution device, video specified coordinate overlay software, image transparency modification software and a video playback device. It marks the local coordinate points in the video being sampled by the surveillance camera to ensure that the user can see the specified coordinate position on the display screen during video recording.

[0016] Compared with the existing technology, the present invention calculates and derives the fire coordinate point in the original video, and then loads the video signal of the video camera at the same position on the scene into the device to locate the fire point at the actual scene, thereby achieving high-precision positioning on the actual scene. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Flowchart of software implementation of the present invention.

[0018] Figure 2 This is a schematic diagram of the characteristic frame conversion picture of the present invention Figure 1 .

[0019] Figure 3 This is a schematic diagram of the characteristic frame conversion picture of the present invention Figure 2 .

[0020] Figure 4 This is the effect diagram of the superposition of the captured picture and the online video of the present invention.

[0021] Figure 5 Schematic diagram of the video coordinates on-site projection of the present invention Figure 1 .

[0022] Figure 6 Schematic diagram of the video coordinates on-site projection of the present invention Figure 2 . DETAILED DESCRIPTION

[0023] The present invention will now be further described with reference to the accompanying drawings.

[0024] See also Figures 1 to 6 The present invention provides a method for locating and restoring a fire point on-site using a surveillance video, comprising the following steps:

[0025] S1: Copy the video footage captured by the monitoring equipment around the fire scene to a computer;

[0026] S2: Use flame features or smoke features to search for feature frames in all frames of the video, record the feature frames where flames or smoke first appear, and convert them into images;

[0027] S3: When abnormally bright pixels are found to appear continuously in the image in S2 (the continuity includes time continuity and pixel coordinate connectivity), all previous frames are backtracked and the frame that initially produced the abnormal brightness is found as the initial frame. The coordinates of all pixels in the image in S2 that are related to the leftmost and rightmost sides of the flame are recorded.

[0028] S4: exporting the coordinates of all flames or smoke in the initial frame in S3 and importing them into the on-site positioning device through a computer interface program;

[0029] S5: The coordinates in S4 are superimposed on the corresponding positions of the online video using a special color by a computer, so as to obtain the effect of the coordinates being marked on the video, or the transparency of the image in S2 is modified in real time, so that the modified image is superimposed on the collected online video to form a new online video. The online video observes the special color position in real time, and the camera position is corrected according to the position of an object fixed at the fire scene before and after the fire, so that the modified image and the new online video overlap based on the position of the object;

[0030] S6: Position the camera at the actual scene based on the flame or smoke obstruction position, including the camera position and the corresponding scene position of the coordinate point at a certain elevation, and determine the closest and farthest possible distances between the flame position at the fire scene and the camera;

[0031] S7: Find a location on the actual site corresponding to the coordinate point in S6 and mark it on the plan;

[0032] S8: The maximum range that these points can form in the plan view in S7 is the possible range of the fire point. That is, according to the on-site location map, mark the range of the fire point in the video. The marked range of the fire point is the range formed by the projection of all positions onto the ground.

[0033] The on-site positioning device includes a surveillance camera address resolution device, video specified coordinate overlay software, image transparency modification software and a video playback device. It marks the local coordinate points in the video being sampled by the surveillance camera to ensure that the user can see the specified coordinate position on the display screen during video recording.

[0034] Example:

[0035] The marked position in the video screenshot is restored to the on-site position. The software flow chart of this embodiment is as follows: Figure 1 As shown in FIG, the video A suspected of fire is exported from the hard disk into a video format that can be recognized by the PC; Figure 2 As shown, video A is searched for feature frames in all frames of the video using flame features or smoke features, and the feature frames where flames or smoke first appear are recorded and converted into picture A1, as shown in Figure 3 As shown, the coordinates B and C of the leftmost and rightmost pixels involved in the flame in image A1 are recorded as (X1, Y1) and (X2, Y2) respectively. Figure 1 The device connection method shown is to import the camera video into the PC, such as Figure 4 As shown, use the software to load the coordinates (X1, Y1) and (X2, Y2) into the PC to overlay the designated coordinate points in real time. If it is a non-mainstream camera, you can also modify the transparency of picture A1 and overlay it with the camera video. In the actual scene, according to the objects that still exist at the scene blocked by flames or smoke, find the target on the scene plan based on the camera's distance range, such as Figure 5 From FG to MN; in the field position, mark the corresponding positions of the left pixel and the right pixel respectively, which are B, L, H on the left and C, O, I on the right, as shown in the following figure: Figure 5 Project them onto the ground respectively, such as Figure 5 Points F, J, H and G, K, I on the left side of the middle; then mark these 6 positions on the plane diagram, such as Figure 6 ; The final fire point range is FGJK Figure 6 shown.

[0036] The above describes the embodiments of the present invention, but the scope of the present invention is not limited thereto. Users can make various changes and implement them without departing from the scope of the present invention, but all of them are included in the scope of protection of this patent.

[0037] The present invention comprehensively solves the technical problems that existing fire video investigations are mainly based on experience, there is no accurate pixel-level video monitoring on-site investigation method, and the fire point is located in the video, resulting in low accuracy of on-site positioning. The fire coordinate point in the original video is calculated and derived, and the video signal of the video camera at the same position on the scene is loaded into the device to locate the fire point at the actual scene, thereby achieving high-precision positioning on the actual scene.

Claims

1. A method for locating and restoring a fire point using surveillance video, characterized in that: The following steps are involved: S1: Copy the video footage captured by the monitoring equipment around the fire scene to a computer; S2: Use flame features or smoke features to search for feature frames in all frames of the video, record the feature frames where flames or smoke first appear, and convert them into images; S3: When abnormally bright pixels appear continuously in the image in S2, the continuity includes the continuity of time and the continuity of pixel coordinates, the frame that initially produces the abnormal brightness is found as the initial frame, and the coordinates of all pixels related to the leftmost and rightmost sides of the flame in the image in S2 are recorded; S4: exporting the coordinates of all flames or smoke in the initial frame in S3 and importing them into the on-site positioning device through a computer interface program; S5: Using a computer, the coordinates in S4 are superimposed on the corresponding positions of the online video using special colors to obtain the effect of the coordinates being marked on the video, or the transparency of the image in S2 is modified in real time, thereby superimposing the modified image with the collected online video to form a new online video. The online video observes the special color position in real time, and the camera position is corrected according to the position of an object fixed at the fire scene before and after the fire, so that the modified image and the new online video overlap based on the position of the object; S6: Position the camera at the actual scene based on the flame or smoke obstruction position, including the camera position and the corresponding scene position of the coordinate point at a certain elevation, and determine the closest and farthest possible distances between the flame position at the fire scene and the camera; S7: Find a location on the actual scene corresponding to the coordinate point in S6, mark three locations corresponding to the left pixel point and three locations corresponding to the right pixel point in the location, make projection points on the ground for each location, and mark them in the plane map; S8: The maximum range that can be formed by these points in the plane diagram in S7 is the possible range of the fire point, that is, according to the on-site location map, the range of the fire point in the video is marked, and the marked fire point range is the range formed by the projection of all positions onto the ground.

2. The method for locating and restoring a fire point using surveillance video according to claim 1, characterized in that: The on-site positioning device includes a surveillance camera address parsing device, video specified coordinate overlay software, image transparency modification software and a video playback device, which marks the local coordinate points in the video being sampled by the surveillance camera to ensure that the user can see the specified coordinate position on the display screen during video recording.

Images