CN112818766A - High-altitude parabolic detection alarm method and system based on computer vision - Google Patents
High-altitude parabolic detection alarm method and system based on computer vision Download PDFInfo
- Publication number
- CN112818766A CN112818766A CN202110061520.3A CN202110061520A CN112818766A CN 112818766 A CN112818766 A CN 112818766A CN 202110061520 A CN202110061520 A CN 202110061520A CN 112818766 A CN112818766 A CN 112818766A
- Authority
- CN
- China
- Prior art keywords
- parabolic
- image
- frame
- moving object
- event
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/44—Event detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a high-altitude parabolic detection alarm method and system based on computer vision, wherein the method comprises the following steps: acquiring a video frame of a monitored object region, converting a gray level image of the video frame, and processing each gray level image by using a frame difference method to obtain a corresponding binary image; extracting a binary image of a continuous multi-frame gray image, comparing the distribution rule characteristics of parabolic events of each pixel point with those of a binary image of an interference event, removing the pixel points corresponding to the interference event, and obtaining a suspected moving object target region in the video image without the interference event; and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment. By using the high-altitude parabolic detection and alarm method based on computer vision, the high-altitude parabolic condition is found in time, the key role in early warning and evidence obtaining of the high-altitude parabolic risk is played, and the life and property safety of people is greatly guaranteed.
Description
Technical Field
The invention relates to the field of video security monitoring, in particular to a high-altitude parabolic detection alarm method and system based on computer vision.
Background
At present, the high-altitude parabolic object has huge potential safety hazard, if the object hits the pedestrian, serious personal injury accidents can be caused, meanwhile, the pedestrian can not be avoided basically due to the conditions of high speed, paroxysmal performance and the like of the high-altitude parabolic object, and the object is difficult to trace back afterwards. With the development of society and the improvement of public security awareness, various measures for dealing with the high-altitude parabolic events are implemented, and the requirements of accurate identification, timely early warning and convenient evidence obtaining are provided for monitoring parabolic event parabolic detection products, so that a high-altitude parabolic detection alarm method and a high-altitude parabolic detection alarm system based on computer vision are needed for realizing real-time high-efficiency intelligent parabolic identification and automatic early warning.
Disclosure of Invention
The embodiment of the invention provides a high-altitude parabolic detection alarm method and system based on computer vision, which can solve the technical problems that parabolic detection monitoring mainly depends on manual work, the efficiency is low, early warning is not timely, information evidence obtaining is insufficient, and the accuracy recognition degree is low in the prior art.
To achieve the above object, a first aspect of the present invention provides a high altitude parabolic detection alarm method based on computer vision, the method comprising: acquiring a video frame for monitoring a parabolic area, and performing gray-scale image conversion on the video frame to obtain a gray-scale image video frame for detecting a parabolic event;
processing each frame of gray level image by using a frame difference method to obtain a corresponding binary image;
extracting a binary image of a continuous multi-frame gray image, comparing the distribution rule characteristics of parabolic events of each pixel point with those of a binary image of an interference event, removing the pixel points corresponding to the interference event, and obtaining a suspected moving object target region in the video image without the interference event;
and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment.
Further, the extracting of the binary image of the continuous multi-frame gray scale image is performed to compare the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, the pixel point corresponding to the interference event is removed, a suspected moving object target area in the video image with the interference event removed is obtained, the storing and counting of the pixel value of each pixel point in the adjacent M frames of images are further included, M is an even number, and the binary image pixel value judgment is performed to obtain a suspected moving object target area result.
Further, the extracting of the binary image of the continuous multi-frame gray level image to compare the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, and removing the pixel point corresponding to the interference event to obtain a suspected moving object target region in the video image from which the interference event is removed further includes: and when the input frame is the ith frame, taking the differential image of the ith-M/2 frame as an image frame for acquiring a moving object, wherein M is an even number, and judging and removing points of motion conditions of the previous frame and the next frame before removal of the differential image of the ith-M/2 frame according to the noise frequency characteristics of the M +1 images.
Further, after removing the point of motion condition of the previous and subsequent frames of the difference image of the i-M/2 th frame, the method further comprises: and extracting the position value of the connected domain of the differential image of the K frames after the i-M/2 to obtain the suspected parabolic track of the continuous K frames.
Further, after the suspected parabolic tracks of the continuous K frames are obtained, all coordinate points are integrated in the same coordinate point set to perform RANSAC algorithm parabolic track fitting judgment.
Further, in the process of performing RANSAC algorithm parabolic trajectory fitting judgment on the suspected moving object, if the result is consistent, decision early warning is performed, including informing a manager of a parabolic event, and storing parabolic point evidence.
In order to achieve the above object, an embodiment of the present invention further provides a high altitude parabolic detection alarm system based on computer vision, which includes an acquisition detection module, a binary image processing module, a moving object target frame difference determination module, a parabolic track fitting early warning module,
the acquisition detection module is used for acquiring a video frame of a monitored parabolic area and converting a gray image of the video frame to obtain a gray image video frame for detecting a parabolic event;
the binary image processing module is used for processing each frame of gray image by using a frame difference method to obtain a corresponding binary image;
the moving object target frame difference judging module is used for extracting a binary image of continuous multi-frame gray level images, comparing the distribution rule characteristics of parabolic events and interference event binary images of each pixel point, removing the pixel points corresponding to the interference events and obtaining a suspected moving object target area in the video image without the interference events;
and the parabolic track fitting early warning module is used for carrying out RANSAC algorithm parabolic track fitting judgment on a suspected moving object, and carrying out decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment.
The embodiment of the invention provides a high-altitude parabolic detection alarm method based on computer vision, which comprises the following steps: acquiring a video frame for monitoring a parabolic area, and performing gray-scale image conversion on the video frame to obtain a gray-scale image video frame for detecting a parabolic event; processing each frame of gray level image by using a frame difference method to obtain a corresponding binary image, extracting the binary images of continuous multi-frame gray level images to compare the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, removing the pixel point corresponding to the interference event, and obtaining a suspected moving object target region in the video image without the interference event; and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment. By using the high-altitude parabolic detection alarm method based on computer vision, each frame of gray level image is processed by using a frame difference method to obtain a corresponding binary image, the binary images of continuous multi-frame gray level images are extracted to compare the distribution rule characteristics of parabolic events and interference event binary images of each pixel point, the pixel points corresponding to interference events are removed, and a suspected moving object target area in a video image without the interference events is obtained; and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment. The risk of throwing objects is accurately identified, early warning is carried out in time, evidence is conveniently obtained, the high-altitude throwing object condition is found in time, evidence is stored by the automatic alarm rear-end platform, the high-altitude throwing object risk early warning and evidence obtaining play a key role, and the life and property safety of people is greatly guaranteed.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart illustrating steps of a high altitude parabolic detection alarm method based on computer vision according to a first embodiment of the present invention;
FIG. 2 is a schematic diagram of a high altitude parabolic detection alarm system based on computer vision according to a second embodiment of the present invention;
FIG. 3 is a diagram illustrating a combination of a plurality of binary images for comparison according to a first embodiment of the present invention;
fig. 4a to 4d are schematic diagrams illustrating a pixel value statistics at a pixel point in a plurality of consecutive binary images according to the first embodiment of the invention.
Detailed Description
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a high altitude parabolic detection alarm method based on computer vision according to a first embodiment of the present invention, in which the method includes:
102, processing each frame of gray level image by using a frame difference method to obtain a corresponding binary image;
103, extracting a binary image of a continuous multi-frame gray image, comparing the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, removing the pixel points corresponding to the interference event, and obtaining a suspected moving object target area in the video image without the interference event; specifically, setting the current frame as the ith frame, and storing all differential binary images from the ith frame to the ith frame by using a container, wherein the total frame is M +1, M is an even number, then the binary image container can sequentially store 7 frames of images in a first-in first-out mode, and when M is 6, the frame images are as shown in FIG. 3;
for each pixel point position, counting the pixel value of each pixel point position in the 7 frames of images, so as to obtain the statistical data of each pixel point position, assuming that i is 7, the pixel position is (x1, y1), and the statistical result of the point to the 7 frames of images is shown in fig. 4a-4 d;
in fig. 4a-4d it is shown that in the (x1, y1) position, the value of the 4 th frame is 255 and the values of the remaining frames are 0, and since the picture has been processed into a binary picture, there are only two possible values of 255 or 0.
For the object which has obtained the suspected motion in the frame, the step removes the non-parabolic motion condition according to some interference factors and the characteristics of the parabolic event. The interference factors have the following characteristics: similar to the situation of leaves and clothes shaking, the occurrence frequency of the motion of the pixel position is higher, and the situation in the statistical chart is represented by the existence of some interference factors and the parabolic event characteristic chart as shown in fig. 4 d:
there are some disturbing factors and the characteristic map of the parabolic event will intermittently appear at pixel values of 255. In case of a parabolic event, when a parabolic object passes through, the object is detected as a moving object, but both the frames before and after the parabolic object appears at the same pixel position are non-moving and can be shaped as "flash", which is shown in the statistical chart as shown in fig. 4a-4c,
continuous x frames in the middle of a characteristic diagram with some interference factors and parabolic events are 255, and the front and rear frames are 0 (x is assumed to be 3 at first), and the judgment is carried out through the noise frequency characteristic of the interference factors, namely, 255 always appears in a high frequency of 'flash' in the characteristic diagram, so that whether the characteristic is noise or not is distinguished.
When the current input frame is the ith frame, the difference image of the ith-M/2 frame (the value is the 4 th frame) is used as the image frame for acquiring the moving object (so that the motion situation of the position in the subsequent time after the suspected moving object appears can be monitored). For the difference image of the i-M/2 th frame, points with the shapes of some interference factors and motion situations of the front and rear frames in the characteristic image of the parabolic event are removed, most of the interference situations can be removed through judgment of noise frequency characteristics of the interference factors, and meanwhile, the trace points of the parabolic event are stored.
103, performing RANSAC algorithm parabolic track fitting judgment on a suspected moving object, performing decision early warning if the suspected moving object is consistent with the RANSAC algorithm parabolic track fitting judgment, specifically, performing noise elimination on the difference image of the i-M/2 th frame, extracting a connected domain of the image, and taking the gravity center of the connected domain as the position value of the moving object. By analogy, each moving object position of each frame of the difference map is saved in a list: [ [ (x11, y11), (x12, y12) ] [ (x21, y21) ]. [ ] ] position points can be extracted, the frame difference map is indicated to exist in a suspected motion point, and if the motion points exist in the continuous K frames, the motion points in the group are suspected parabolic tracks. If a suspected parabolic track exists, all coordinate points of the suspected parabolic track are integrated into the same coordinate point set, whether the suspected parabolic track conforms to the situation of the parabolic track is judged by using a RANSAC algorithm, and meanwhile, part of noise can be removed. Ransac (random SAmple consensus) is a random sampling consensus algorithm, and since a curve which is most consistent with a general parabolic trajectory is a parabola in a quadratic curve, quadratic equation fitting is adopted for the parabolic trajectory fitting algorithm in the application.
The method also comprises the following subdivision steps:
further, the extracting of the binary image of the continuous multi-frame gray scale image is performed to compare the distribution rule characteristics of the parabolic event of each pixel point with the binary image of the interference event, the pixel point corresponding to the interference event is removed, a suspected moving object target region in the video image from which the interference event is removed is obtained, the storing and counting of pixel values of each pixel point in adjacent M frames of images is further included, M is an even number, the binary image pixel value judgment is performed, a suspected moving object target region result is obtained, that is, the current frame is set as the ith frame, and a container is used to store all binary images from the ith frame to the ith frame i, wherein M is an even number, and since the image is processed into the binary image, only two possible values of 255 or 0 exist, and the obtained suspected moving object region has a value of 255. Further, the extracting of the binary image of the continuous multi-frame grayscale image is performed to compare the distribution rule characteristics of the parabolic event of each pixel point with the binary image of the interference event, and the removing of the pixel points corresponding to the interference event is performed to obtain the suspected moving object target region in the video image from which the interference event is removed, and the obtaining of the suspected moving object target region further includes that when the input frame is the i-th frame, the differential image of the i-M/2-th frame is used to obtain the image frame of the moving object, M is an even number, and the points of the differential image of the i-M/2-th frame where the motion conditions of the previous and subsequent frames occur are removed are judged through the noise frequency characteristics of the M +1 images, so that most of the noise is removed, and the step 102 has already explained the non-parabolic motion. The interference factors have the following characteristics: similar to the situation of leaves and clothes shaking, the motion of the pixel position occurs more frequently.
Further, after removing the point of motion condition of the previous and subsequent frames of the difference image of the i-M/2 th frame, the method further comprises: and (4) extracting a connected domain position value of the differential image of the K frames after the (i-M/2) th frame to obtain a suspected parabolic track of the continuous K frames, namely extracting a gravity center position value of a connected domain of the parabolic track.
Further, after the suspected parabolic tracks of the continuous K frames are obtained, all coordinate points are integrated in the same coordinate point set to perform RANSAC algorithm parabolic track fitting judgment, and the process is as follows:
1. randomly selecting a plurality of coordinate points, and calculating a quadratic fitting curve equation model M;
2. calculating all coordinate points of the point set, and adding an inner point set F if the projection error in the model M is smaller than a threshold value;
3. if the number of the coordinate points of the inner point set in the step 2 is larger than that of the current optimal inner point set F _ best, updating the F _ best, and accumulating the iteration times i;
4. and if the iteration number i is larger than the threshold value, stopping iteration, and obtaining a fitted curve corresponding to the F _ best, namely a fitted parabolic curve. Otherwise, the iteration is continued.
5. If the number of the inner point set F _ best is larger than a threshold value K1 (0 < K1< K in principle), the inner point set F _ best is defined as a final parabolic track, the corresponding inner point is defined as a track point, and the outer point is discarded (defined as noise); otherwise, the parabolic curve and all coordinate points are discarded.
Through the steps, whether the event continuously having the suspected motion points of the K frames is a parabolic event or not can be judged. Further, performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, if the suspected moving object is matched with the RANSAC algorithm parabolic track fitting judgment, performing decision early warning including informing a manager of parabolic event existence, storing parabolic point evidence, and optionally performing picture output of track points through a monitoring screen to allow a worker to confirm the parabolic throwing position; or optionally notifying the manager of the occurrence of the parabolic event by means of short messages and the like.
As shown in fig. 2, an embodiment of the present invention further provides a high-altitude parabolic detection alarm system based on computer vision, which includes an acquisition detection module, a binary image processing module, a moving object target frame difference determination module, a parabolic track fitting early warning module,
the acquisition detection module is used for acquiring video frames of the monitored parabolic region and converting a gray image to obtain a gray image video frame for detecting a parabolic event;
the binary image processing module is used for processing each frame of gray image by using a frame difference method to obtain a corresponding binary image; the moving object target frame difference judging module is used for extracting a binary image of continuous multi-frame gray level images, comparing the distribution rule characteristics of parabolic events and interference event binary images of each pixel point, removing the pixel points corresponding to the interference events and obtaining a suspected moving object target area in the video image without the interference events;
and the parabolic track fitting early warning module is used for carrying out RANSAC algorithm parabolic track fitting judgment on a suspected moving object, and carrying out decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment.
Also included is an apparatus for high altitude parabolic detection alert based on computer vision, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the method when executing the computer program.
The above are all system modules in the above method, and the description of the embodiments is not repeated here.
The invention has the beneficial technical effects that: the embodiment of the invention provides a high-altitude parabolic detection alarm method based on computer vision, which comprises the following steps: acquiring a video frame of a monitored object region, and converting a gray-scale image to obtain a gray-scale image video frame for detecting a parabolic event; processing each frame of gray level image by using a frame difference method to obtain a corresponding binary image, extracting the binary images of continuous multi-frame gray level images to compare the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, removing the pixel point corresponding to the interference event, and obtaining a suspected moving object target region in the video image without the interference event; and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment. The high-altitude parabolic detection alarm method based on computer vision is applied, each frame gray level image is processed by using a frame difference method to obtain a corresponding binary image, the binary images of continuous multi-frame gray level images are extracted to compare the distribution rule characteristics of parabolic events and interference event binary images of each pixel point, the pixel points corresponding to interference events are removed, a suspected moving object target area in a video image with the interference events removed is obtained, and a suspected moving object target area in the video image with noise points removed is obtained; and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment. The risk of throwing objects is accurately identified, early warning is carried out in time, evidence is conveniently obtained, the high-altitude throwing object condition is found in time, evidence is stored by the automatic alarm rear-end platform, the high-altitude throwing object risk early warning and evidence obtaining play a key role, and the life and property safety of people is greatly guaranteed. Meanwhile, in the monitoring area, the system can detect all suspected moving targets in the video frame by using a frame difference method, and can also detect relatively small moving targets, so that the detection specification of the parabolic detection system is ensured. Meanwhile, the frame difference needs a small amount of calculation, and the overall detection speed can be ensured. Each iteration uses the differential image of the (i-M/2) th frame as a binary image for extracting a connected domain of a motion region, instead of using the differential image of the current frame (i-th frame). The differential image of the i-M/2 th frame can simultaneously use the differential images in front of and behind the frame image to filter the moving target, and can better filter noise.
By counting the distribution of pixel values of the same coordinate point in a plurality of frames, the condition that a statistical graph with high-frequency distribution is noise can be distinguished, and various noises such as leaves, clothes and the like are removed according to the condition. The method performs parabolic track fitting when the moving target appears in a plurality of continuous frames of the differential image filtered by the moving target noise, can better filter the isolated noise which happens in a time domain in advance, and improves the overall efficiency. And for a difference image set with a plurality of continuous moving targets, fitting and judging the coordinate point set by using a RANSAC algorithm, and extracting a quadratic equation curve of the coordinate point set with the maximum number of points as a parabolic trajectory line, so that an actual parabola can be better fitted. The RANSAC algorithm can effectively obtain the result of distinguishing the relatively most reasonable inner point from the outer point in the coordinate point set, so that the noise (namely the outer point) in the difference image can be further removed.
It should be noted that, for the sake of simplicity, the above-mentioned method embodiments are described as a series of acts or combinations, but those skilled in the art should understand that the present invention is not limited by the described order of acts, as some steps may be performed in other orders or simultaneously according to the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no acts or modules are necessarily required of the invention.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
The foregoing is a description of the present invention, and those skilled in the art will recognize that there are many variations in the form and details of the embodiments and applications of the present invention, and that the present invention is not limited by the details of the embodiments and applications illustrated in the accompanying drawings.
Claims (8)
1. A high altitude parabolic detection alarm method based on computer vision, the method comprising:
acquiring a video frame for monitoring a parabolic area, and performing gray-scale image conversion on the video frame to obtain a gray-scale image video frame for detecting a parabolic event;
processing each frame of gray level image by using a frame difference method to obtain a corresponding binary image;
extracting a binary image of a continuous multi-frame gray image, comparing the distribution rule characteristics of parabolic events of each pixel point with those of a binary image of an interference event, removing the pixel points corresponding to the interference event, and obtaining a suspected moving object target region in the video image without the interference event;
and (4) performing RANSAC algorithm parabolic track fitting judgment on the suspected moving object, and performing decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment.
2. The method of claim 1, wherein the extracting the binary image of the continuous multi-frame gray scale image to compare the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, and removing the pixel point corresponding to the interference event to obtain the target region of the suspected moving object in the video image without the interference event comprises: and counting and storing the pixel value of each pixel point in the adjacent M frames of images, wherein M is an even number, and judging the binary image pixel value to obtain a suspected moving object target area result.
3. The method of claim 2, wherein the extracting the binary image of the continuous multi-frame gray scale image to compare the distribution rule characteristics of the parabolic event and the binary image of the interference event of each pixel point, and removing the pixel point corresponding to the interference event to obtain the target region of the suspected moving object in the video image without the interference event further comprises:
and when the input frame is the ith frame, taking the differential image of the ith-M/2 frame as an image frame for acquiring a moving object, and judging and removing points of motion conditions of the frames before and after the differential image of the ith-M/2 frame according to the noise frequency characteristics of the M +1 images.
4. The method of detecting an alarm according to claim 3, wherein said removing the points of motion occurrence in the previous and subsequent frames of the difference image of the i-M/2 th frame further comprises: and extracting the position value of the connected domain of the differential image of the K frames after the i-M/2 to obtain the suspected parabolic track of the continuous K frames.
5. The detection alarm method according to claim 4, wherein after the suspected parabolic trajectory of the continuous K frames is obtained, all coordinate points are integrated in the same coordinate point set to perform RANSAC algorithm parabolic trajectory fitting judgment.
6. The detection alarm method according to claim 1, wherein in the process of performing RANSAC algorithm parabolic trajectory fitting judgment on the suspected moving object, if the suspected moving object is matched with the RANSAC algorithm parabolic trajectory fitting judgment, decision early warning is performed, a manager is notified to be reminded of a parabolic event, and parabolic point evidence is stored.
7. A high-altitude parabolic detection alarm system based on computer vision is characterized by comprising an acquisition detection module, a binary image processing module, a moving object target frame difference judgment module and a parabolic track fitting early warning module,
the acquisition detection module is used for acquiring a video frame of a monitored parabolic area and converting a gray image of the video frame to obtain a gray image video frame for detecting a parabolic event;
the binary image processing module is used for processing each frame of gray image by using a frame difference method to obtain a corresponding binary image;
the moving object target frame difference judging module is used for extracting a binary image of continuous multi-frame gray level images, comparing the distribution rule characteristics of parabolic events and interference event binary images of each pixel point, removing the pixel points corresponding to the interference events and obtaining a suspected moving object target area in the video image without the interference events;
and the parabolic track fitting early warning module is used for carrying out RANSAC algorithm parabolic track fitting judgment on a suspected moving object, and carrying out decision early warning if the suspected moving object is in accordance with the RANSAC algorithm parabolic track fitting judgment.
8. An apparatus for high altitude parabolic detection alert based on computer vision, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110061520.3A CN112818766A (en) | 2021-01-18 | 2021-01-18 | High-altitude parabolic detection alarm method and system based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110061520.3A CN112818766A (en) | 2021-01-18 | 2021-01-18 | High-altitude parabolic detection alarm method and system based on computer vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112818766A true CN112818766A (en) | 2021-05-18 |
Family
ID=75869861
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110061520.3A Pending CN112818766A (en) | 2021-01-18 | 2021-01-18 | High-altitude parabolic detection alarm method and system based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112818766A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113255697A (en) * | 2021-06-01 | 2021-08-13 | 南京图菱视频科技有限公司 | High-precision high-altitude parabolic detection system and method under complex scene |
| CN114639172A (en) * | 2022-05-18 | 2022-06-17 | 合肥的卢深视科技有限公司 | High-altitude parabolic early warning method and system, electronic equipment and storage medium |
| CN114693556A (en) * | 2022-03-25 | 2022-07-01 | 英特灵达信息技术(深圳)有限公司 | Method for detecting and removing smear of moving target by high-altitude parabolic frame difference method |
| CN115719362A (en) * | 2022-05-31 | 2023-02-28 | 海纳云物联科技有限公司 | High-altitude parabolic detection method, system, equipment and storage medium |
| CN117237676A (en) * | 2023-11-09 | 2023-12-15 | 中核国电漳州能源有限公司 | Method for processing small target drop track of nuclear power plant based on event camera |
-
2021
- 2021-01-18 CN CN202110061520.3A patent/CN112818766A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113255697A (en) * | 2021-06-01 | 2021-08-13 | 南京图菱视频科技有限公司 | High-precision high-altitude parabolic detection system and method under complex scene |
| CN114693556A (en) * | 2022-03-25 | 2022-07-01 | 英特灵达信息技术(深圳)有限公司 | Method for detecting and removing smear of moving target by high-altitude parabolic frame difference method |
| CN114639172A (en) * | 2022-05-18 | 2022-06-17 | 合肥的卢深视科技有限公司 | High-altitude parabolic early warning method and system, electronic equipment and storage medium |
| CN115719362A (en) * | 2022-05-31 | 2023-02-28 | 海纳云物联科技有限公司 | High-altitude parabolic detection method, system, equipment and storage medium |
| CN117237676A (en) * | 2023-11-09 | 2023-12-15 | 中核国电漳州能源有限公司 | Method for processing small target drop track of nuclear power plant based on event camera |
| CN117237676B (en) * | 2023-11-09 | 2024-03-01 | 中核国电漳州能源有限公司 | Method for processing small target drop track of nuclear power plant based on event camera |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112818766A (en) | High-altitude parabolic detection alarm method and system based on computer vision | |
| CN107527009B (en) | Remnant detection method based on YOLO target detection | |
| CN101266689B (en) | A mobile target detection method and device | |
| CN102073851A (en) | Method and system for automatically identifying urban traffic accident | |
| CN102831442A (en) | Abnormal behavior detection method and equipment and method and equipment for generating abnormal behavior detection equipment | |
| CN107944354B (en) | Vehicle detection method based on deep learning | |
| CN111723773B (en) | Method and device for detecting carryover, electronic equipment and readable storage medium | |
| WO2023273010A1 (en) | High-rise littering detection method, apparatus, and device, and computer storage medium | |
| CN113409362B (en) | High altitude parabolic detection method and device, equipment and computer storage medium | |
| Han et al. | Development of early tunnel fire detection algorithm using the image processing | |
| CN110544271B (en) | Parabolic motion detection method and related device | |
| CN105469054A (en) | Model construction method of normal behaviors and detection method of abnormal behaviors | |
| CN110782409A (en) | Method for removing shadow of multi-motion object | |
| CN101324959B (en) | Method and apparatus for detecting moving target | |
| CN113255697B (en) | High-precision high-altitude parabolic detection system and method in complex scene | |
| CN110543827B (en) | Multi-class vehicle detection method based on Gaussian mixture model and deep learning | |
| WO2012081969A1 (en) | A system and method to detect intrusion event | |
| CN108960181B (en) | Black smoke vehicle detection method based on multi-scale block LBP and hidden Markov model | |
| CN112581489A (en) | Video compression method, device and storage medium | |
| CN107403192B (en) | Multi-classifier-based rapid target detection method and system | |
| Chen et al. | Intrusion detection of specific area based on video | |
| JP2024516642A (en) | Behavior detection method, electronic device and computer-readable storage medium | |
| Chen et al. | A hierarchical accident recognition method for highway traffic systems | |
| CN111025288B (en) | Security radar monitoring device and system | |
| Frías-Velázquez et al. | A mathematical morphology-based approach for vehicle detection in road tunnels |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |