WO2022001961A1 - 一种高空抛物动目标检测方法、检测设备和检测系统 - Google Patents

一种高空抛物动目标检测方法、检测设备和检测系统 Download PDF

Info

Publication number
WO2022001961A1
WO2022001961A1 PCT/CN2021/102769 CN2021102769W WO2022001961A1 WO 2022001961 A1 WO2022001961 A1 WO 2022001961A1 CN 2021102769 W CN2021102769 W CN 2021102769W WO 2022001961 A1 WO2022001961 A1 WO 2022001961A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
images
difference
moving target
differential
Prior art date
Application number
PCT/CN2021/102769
Other languages
English (en)
French (fr)
Inventor
周飞
刘倞
Original Assignee
深圳天感智能有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 深圳天感智能有限公司 filed Critical 深圳天感智能有限公司
Publication of WO2022001961A1 publication Critical patent/WO2022001961A1/zh

Links

Images

Classifications

    • 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
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • 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
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20224—Image subtraction
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30232—Surveillance

Definitions

  • the present application relates to the field of monitoring, and in particular, to a detection method, detection equipment and detection system for high-altitude parabolic moving targets.
  • the traditional monitoring method is to arrange for a special person to be on duty.
  • the method of arranging a special person to be on duty is expensive, and most falling objects are relatively small, and the human eye may not be able to track where the fallen objects were thrown from.
  • an embodiment of the present application provides a method for detecting a high-altitude parabolic moving target, the method comprising:
  • the connected domain analysis is performed on the morphological result image to determine the position of the high-altitude parabolic moving target.
  • the obtaining the preprocessed image and performing the difference operation between frames, and obtaining the difference result image includes:
  • a difference result image is generated according to the difference result.
  • the obtaining the preprocessed image and performing the difference operation between frames, and obtaining the difference result image includes:
  • a difference result image is generated according to the difference result.
  • the obtaining the extreme value of the pixel value at the same position of the Y difference images as the difference result includes:
  • the performing connected domain analysis on the morphological result image to determine the location of the high-altitude parabolic moving target includes:
  • the center coordinates are taken as the location of the high-altitude parabolic object.
  • the embodiments of the present application also provide a high-altitude parabolic moving target detection device, including:
  • the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the above-described method for detecting a high-altitude parabolic moving target.
  • an embodiment of the present application further provides a high-altitude parabolic moving target detection system, including the above-mentioned high-altitude parabolic moving target detection device and image acquisition device;
  • the image acquisition device is connected to the high-altitude parabolic moving target detection device, and the high-altitude parabolic moving target detection device is configured to receive and process images sent by the image acquisition device.
  • the image acquisition device is an infrared signal acquisition device.
  • the infrared signal acquisition device includes a combination of a star-level camera and an infrared fill light, or at least one of an infrared camera.
  • embodiments of the present application further provide a non-volatile computer-readable storage medium, where the computer-readable storage medium stores computer-executable instructions, when the computer-executable instructions are executed by a processor , causing the processor to execute the above-mentioned high-altitude parabolic moving target detection method.
  • the beneficial effects of the present application are: different from the prior art, the high-altitude parabolic moving target detection method, the detection device and the detection system in the embodiments of the present application acquire several consecutive images and analyze all the objects. Perform preprocessing on several consecutive images to obtain a preprocessed image, then perform an inter-frame difference operation on the preprocessed image to obtain a difference result image, and then perform image processing on the difference result image to obtain a binarized image, and further , perform morphological operations on the binarized image to obtain a morphological result image, and finally perform a connected domain analysis on the morphological result image, not only to identify the high-altitude parabolic moving target, but also to determine the location of the high-altitude parabolic moving target.
  • FIG. 1 is a schematic diagram of an application scenario of a method for detecting high-altitude parabolic moving targets in an embodiment of the present application
  • Fig. 2 is a flow chart of a method for detecting high-altitude parabolic moving targets in an embodiment of the present application
  • FIG. 3 is a flowchart of obtaining a differential result image in an embodiment of the present application.
  • FIG. 4 is a schematic structural diagram of a high-altitude parabolic moving target detection device in an embodiment of the present application
  • FIG. 5 is a schematic diagram of the hardware structure of a high-altitude parabolic moving target detection device in an embodiment of the present application.
  • the high-altitude parabolic moving target detection method provided by the embodiment of the present application is suitable for the application scenario shown in FIG. 1 .
  • the application scenario is a high-altitude parabolic moving target detection system, including high-altitude parabolic moving target detection equipment and image acquisition.
  • a device, the image acquisition device is communicatively connected to the high-altitude parabolic moving target detection device, for example, it can be connected through a network cable.
  • the system further includes a switch.
  • the image acquisition device may also be connected to the high-altitude parabolic moving target detection device through a switch.
  • Fig. 1 exemplarily shows that the image capturing device 20A, the image capturing device 20B and the image capturing device 20C are connected to the high-altitude parabolic moving target detection device 10 through the switch 30, respectively.
  • the image acquisition device 20 is used for sending the real-time acquired images to the high-altitude parabolic moving target detection device 10 for processing and storage.
  • the image acquisition device 20 may be any infrared signal acquisition device capable of detecting thermal infrared signals within the sensing range, for example, a combination of a star-level camera and an infrared fill light, or an infrared camera.
  • the methods provided in the embodiments of the present application can be further extended to other suitable application environments, and are not limited to the application environment shown in FIG. 1 .
  • the application environment may also include More or less image acquisition equipment, high-altitude parabolic moving target detection equipment, and switches.
  • the high-altitude parabolic moving target detection method provided by the embodiments of the present application can monitor and identify the method and target position of high-altitude parabolic events.
  • the specific types of moving targets are not limited, but commonly identified are water bottles, cigarette butts, and other household garbage.
  • the moving target detection algorithms include optical flow method, background subtraction method and inter-frame difference method.
  • the optical flow method detects moving objects, and its basic idea is to assign a velocity vector to each pixel in the image, thereby forming the motion field of the image.
  • the image is dynamically analyzed according to the velocity vector characteristics of each pixel. If there is no moving target in the image, the optical flow vector changes continuously in the entire image area, and when there is relative motion between the object and the image background, the velocity vector formed by the moving object must be different from the velocity of the neighborhood background. vector to detect the position of the moving object.
  • the detection time limit of the optical flow method can accurately calculate the speed of the moving object without knowing any information of the scene. It can not only be applied to the detection of moving objects in the static background, but also can be used to detect moving objects in the dynamic background when the camera is moving. Since the optical flow method needs to perform iterative operations, the higher the precision, the greater the amount of computation. Therefore, the optical flow method is a time-consuming algorithm, and it is difficult to meet the real-time requirements. In addition, the anti-noise performance of the optical flow method is relatively poor. When the illumination changes, even if there is no movement, the optical flow will exist, which is prone to false detection.
  • the background subtraction method is a common moving object detection algorithm.
  • the basic idea is to use the parameter model of the background to approximate the pixel value of the background image, and compare the current frame with the background image to detect the moving area.
  • the pixel area is considered as the motion area, and the pixel area with less distinction is considered as the background area.
  • the background subtraction method must have a background image, and the background image must be updated in real time with the change of the illumination or the external environment, so the key of the background subtraction method is the background modeling and its update. Under conditions such as weather changes, rain and reflections, it is easy to identify the background as the foreground using the background subtraction method.
  • the inter-frame difference method is one of the most commonly used moving object detection and segmentation methods.
  • the background does not accumulate, and the update speed is fast, the algorithm is simple, and the calculation amount is small, which is suitable for scenarios with high real-time requirements.
  • the disadvantage of the algorithm is that it is more sensitive to environmental noise, and the selection of the threshold is very critical. If the selection is too low, it is not enough to suppress the noise in the image, and if it is too high, the useful changes in the image are ignored. For a relatively large moving target with the same color, there may be holes inside the target, and the moving target cannot be completely extracted.
  • an embodiment of the present application provides a method for detecting a high-altitude parabolic moving target.
  • the method is performed by a high-altitude parabolic moving target detection device, and the method includes:
  • Step 202 acquiring several consecutive images.
  • the image acquisition device is placed in the area that needs to be monitored, and image acquisition is performed on the monitored area within 24 hours.
  • the high-altitude parabolic moving target detection equipment acquires the monitoring images collected by the image acquisition equipment in real time, and converts the continuous images into videos for storage, which is convenient for subsequent tracking.
  • the time interval of several consecutive images acquired is 1 minute. It can be understood that, in some other embodiments, the time interval of several consecutive images obtained can be set by itself according to the actual situation, and it is not necessary to be bound by the limitation in this embodiment.
  • Step 204 Perform image preprocessing on several of the consecutive images.
  • image preprocessing includes image color space conversion and image filtering.
  • the acquired image is an RGB image
  • the RGB image is converted into a grayscale image by a specific algorithm, for example, the RGB image is converted into a grayscale image by a maximum value method, an average value method, or a weighted average method.
  • the grayscale image is further filtered, so that the noise in the image can be removed, which is beneficial to the subsequent inter-frame difference operation and image processing.
  • the grayscale image may be subjected to mean filtering, or the grayscale image may be subjected to operations such as median filtering, Gaussian filtering, or bilateral filtering, which can effectively remove noise in the image background.
  • the acquired RGB image can also be converted into an HSV image, and then the color channel separation is performed on the HSV image, and the H channel is taken as the detection channel, so as to obtain a grayscale image, and then The grayscale image is filtered according to the method described above, and the noise in the background of the image can also be removed.
  • Step 206 Obtain the preprocessed image and perform an inter-frame difference operation to obtain a difference result image.
  • the inter-frame difference operation is an improved image inter-frame difference method, and the method for calculating the inter-frame difference is to perform a difference calculation after shifting the image, and after multiple translations, the difference is obtained, and the pole of the absolute value of the difference is taken. value as the final difference result.
  • the difference image is obtained by directly taking two frames of consecutive images to obtain the difference image
  • the three-frame difference method takes the preprocessed image of three consecutive frames, and differentiates the previous frame image and the intermediate frame image to obtain the difference image 1
  • the difference image 2 is obtained by the difference between the previous frame image and the intermediate frame image
  • the difference image 3 is obtained from the difference image 1 and the image 2
  • the difference image 3 is used as the difference result image.
  • the inter-frame difference method in the embodiment of the present application filters out most of the interference caused by environmental factors such as camera shake, rain, trees, and clothing by means of translation and re-differentiation, and not only improves the traditional two-frame method and three-frame method.
  • the disadvantage is that it is sensitive to environmental noise, and it can completely extract moving objects.
  • the image difference is performed on the preprocessed image by the improved image frame difference method, and the step of obtaining the difference result image may specifically include:
  • Step 302 Acquire two consecutive frames of images from the preprocessed images to obtain a previous frame image and a subsequent frame image.
  • the previous frame image is the previous frame grayscale image
  • the latter frame image is the latter frame grayscale image.
  • the high-altitude parabolic moving target detection device obtains two consecutive grayscale images from the filtered grayscale images, and obtains the previous frame grayscale image and the latter frame grayscale image.
  • Step 304 Translate the previous frame image along the vector (M, N) to obtain Z translation images, where M ⁇ [-m,m], N ⁇ [-n,n], M, N are integers, Z is an integer greater than or equal to 1, and m and n are integers.
  • the previous frame grayscale image is translated along the vector (M, N) to obtain Z translation images, and the Z translation images are specifically (2m+1) ⁇ (2n+1).
  • M ⁇ [-m,m], N ⁇ [-n,n], M, N are integers
  • Z is an integer greater than or equal to 1
  • m and n are integers
  • the values of m and n can be greater than or equal to 0 and an integer less than the width and height of the image.
  • the image of the previous frame and the image of the subsequent frame can be interchanged, and the object to be translated can also be the image of the subsequent frame.
  • the following frame images are translated along the vector (M, N) to obtain Z translation images, where M ⁇ [-m,m], N ⁇ [-n,n], M, N are integers, and Z is An integer greater than or equal to 1, where m and n are integers.
  • Step 306 Perform a difference operation on the Z translation images and the subsequent frame images to obtain Y difference images, where Y is an integer.
  • the difference image is also a frame difference image.
  • the difference operation is performed on Z translation images, that is, 441 translation images, and the grayscale images of the subsequent frames, that is, frame difference is performed, and 441 frame difference images are obtained.
  • a difference operation is performed between the Z translation images and the previous frame image to obtain Y difference images, where Y is an integer.
  • Step 308 Obtain the extreme value of the pixel value at the same position of the Y difference images as the difference result.
  • the difference result is obtained by obtaining the extreme value of the pixel value at the same position of the Y difference images.
  • the pixel values at the same positions in the Y differential images can be obtained, and then the magnitudes of the values at each same pixel position in the Y differential images can be compared, and the maximum value can be taken as the differential result.
  • the minimum value of the pixel values at the same position can also be taken as the difference result.
  • Step 310 Generate a difference result image according to the difference result.
  • the high-altitude parabolic moving target detection device generates a difference result image according to the difference result, and the value of each pixel in the difference result image is the extreme value of the difference image at the same position.
  • step 208 is executed to perform image processing on the difference result image to obtain a binarized image.
  • the image processing includes image filtering and image segmentation. Specifically, mean filtering is performed on the difference result image, and a local adaptive threshold method operation is performed on the filtered image to obtain a binarized image.
  • image filtering methods such as Gaussian filtering, bilateral filtering, and median filtering may be used to replace the mean filtering in this embodiment, and a fixed threshold method, other automatic filtering methods may be used.
  • the adaptive threshold method, the region growing method and the watershed method replace the local adaptive method of this embodiment.
  • Step 210 Perform morphological operations on the binarized image to obtain a morphological result image.
  • the morphological operation includes one or more of erosion, dilation, opening operation, and closing operation.
  • the morphological result image is the binarized image after dilation. Specifically, a closed operation operation is performed on the binarized image to obtain a morphological result image, that is, an expanded binarized image.
  • Step 212 Perform connected domain analysis on the morphological result image to determine the location of the high-altitude parabolic moving target.
  • the connected domain analysis is performed on the morphological result image, that is, the expanded binarized image, to obtain the contour of the potential target, and then the contour of the potential target is extracted and screened according to the area size, perimeter and heart rate of the contour, Then, the centroid of the contour is obtained by calculating the image moment of the target contour, and the centroid is used as the center coordinate of the contour, and the center coordinate is the position of the moving target, thus the position of the high-altitude parabolic moving target can be determined.
  • a preprocessed image is obtained, and then an inter-frame difference operation is performed on the preprocessed image to obtain a difference result image, and then Perform image processing on the difference result image to obtain a binarized image, further, perform morphological operations on the binarized image to obtain a morphological result image, and finally perform a connected domain analysis on the morphological result image, not only can identify High-altitude parabolic moving target, and can determine the location of high-altitude parabolic moving target.
  • the embodiment of the present application also provides a high-altitude parabolic moving target detection device 400, as shown in FIG. 4, including:
  • a first processing module 404 configured to perform image preprocessing on several of the continuous images
  • the calculation module 406 is used to obtain the preprocessed image and perform the difference operation between frames to obtain the difference result image;
  • the second processing module 408 is configured to perform image processing on the difference result image to obtain a binarized image
  • a third processing module 410 configured to perform morphological operations on the binarized image to obtain a morphological result image
  • the analysis module 412 is configured to perform connected domain analysis on the morphological result image to determine the position of the high-altitude parabolic moving target.
  • the computing module 406 is specifically configured to:
  • a difference result image is generated according to the difference result.
  • a difference result image is generated according to the difference result.
  • the analysis module 412 is specifically configured to:
  • the center coordinates are taken as the location of the high-altitude parabolic object.
  • the high-altitude parabolic moving target detection device acquires several continuous images through the acquisition module, and preprocesses the several continuous images through the first processing module to obtain preprocessed images, and then uses the calculation module to perform preprocessing on the preprocessed images.
  • the processed image is subjected to an inter-frame difference operation to obtain a difference result image, and then image processing is performed on the difference result image by the second processing module to obtain a binarized image, and further, the binarized image is processed by a third processing module.
  • Perform morphological operations to obtain morphological result images, and finally perform connected domain analysis on the morphological result images through the analysis module which can not only identify high-altitude parabolic moving targets, but also determine the location of high-altitude parabolic moving targets.
  • the above-mentioned high-altitude parabolic moving target detection device can execute the high-altitude parabolic moving target detection method provided by the embodiment of the present application, and has functional modules and beneficial effects corresponding to the execution method.
  • the high-altitude parabolic moving target detection apparatus For technical details not described in detail in the embodiments of the high-altitude parabolic moving target detection apparatus, reference may be made to the high-altitude parabolic moving target detection method provided by the embodiments of the present application.
  • FIG. 5 is a schematic diagram of the hardware structure of a high-altitude parabolic moving target detection device provided by an embodiment of the present application. As shown in FIG. 5 , the high-altitude parabolic moving target detection device 500 includes:
  • processors 502 and a memory 504, one processor 502 is taken as an example in FIG. 5 .
  • the processor 502 may be a central processing unit, or an image processor, or the processor 502 may be a processor that integrates the functions of the central processing unit and the image processor.
  • the processor 502 and the memory 504 may be connected by a bus or in other ways, and the connection by a bus is taken as an example in FIG. 5 .
  • the memory 504 can be used to store non-volatile software programs, non-volatile computer-executable programs and modules, such as corresponding to the high-altitude parabolic moving target detection method in the embodiment of the present application. (for example, the acquisition module 402, the first processing module 404, the calculation module 406, the second processing module 408, the third processing module 410, and the analysis module 412 shown in FIG. 4).
  • the processor 502 executes various functional applications and data processing of the high-altitude parabolic moving target detection device by running the non-volatile software programs, instructions and modules stored in the memory 504, that is, to realize the high-altitude parabolic motion in the above method embodiments. object detection method.
  • the memory 504 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the high-altitude parabolic moving target detection device and the like. Additionally, memory 504 may include high speed random access memory, and may also include nonvolatile memory, such as at least one magnetic disk storage device, flash memory device, or other nonvolatile solid state storage device. In some embodiments, the memory 504 may optionally include memory located remotely from the processor 502, and these remote memories may be connected to the high-altitude parabolic moving object detection device via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
  • the one or more modules are stored in the memory 504, and when executed by the one or more high-altitude parabolic moving target detection devices, execute the high-altitude parabolic moving target detection method in any of the above method embodiments, for example, execute The method steps 202 to 212 in FIG. 2 and the method steps 302 to 310 in FIG. 3 described above; implement the functions of the modules 402 to 412 in FIG. 4 .
  • Embodiments of the present application also provide a computer program product, including a computer program stored on a non-volatile computer-readable storage medium, where the computer program includes program instructions, when the program instructions are executed by a computer, The computer executes: method steps 202 to 212 in FIG. 2 , and method steps 302 to 310 in FIG. 3 .
  • the above product can execute the method provided by the embodiments of the present application, and has functional modules and beneficial effects corresponding to the execution method.
  • the methods provided in the embodiments of the present application can execute the method provided by the embodiments of the present application, and has functional modules and beneficial effects corresponding to the execution method.
  • the high-altitude parabolic moving target detection devices in the embodiments of the present application exist in various forms, including but not limited to:
  • Mobile communication equipment This type of equipment is characterized by having mobile communication functions, and its main goal is to provide voice and data communication.
  • Such terminals include: smart phones (eg iPhone), multimedia phones, feature phones, and low-end phones.
  • Ultra-mobile personal computer equipment This type of equipment belongs to the category of personal computers, has computing and processing functions, and generally has the characteristics of mobile Internet access.
  • Such terminals include: PDA, MID and UMPC devices, such as iPad.
  • Portable entertainment equipment This type of equipment can display and play multimedia content.
  • Such devices include: audio and video players (eg iPod), handheld game consoles, e-books, as well as smart toys and portable car navigation devices.
  • the device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
  • each embodiment can be implemented by means of a program plus a general hardware platform, and certainly can also be implemented by hardware.
  • Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and the program is During execution, it may include the processes of the embodiments of the above-mentioned methods.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM) or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

一种高空抛物动目标检测方法、检测设备和检测系统，涉及监控领域，包括获取若干连续图像（202）；对若干所述连续图像进行图像预处理（204）；获取预处理后的图像并进行帧间差分运算，获得差分结果图像（206）；对所述差分结果图像进行图像处理，获得二值化图像（208）；对所述二值化图像进行形态学操作，获得形态学结果图像（210）；对所述形态学结果图像进行连通域分析，确定高空抛物动目标所处位置（212）。

Description

一种高空抛物动目标检测方法、检测设备和检测系统
相关申请交叉引用
本申请要求于2020年06月28日申请的、申请号为202010598843.1，申请名称为“一种高空抛物动目标检测方法、检测设备和检测系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及监控领域，特别是涉及一种高空抛物动目标检测方法、检测设备和检测系统。
背景技术
高空抛物事件在生活中屡屡报导，不断地危害着城市居民明的生命财产安全。但是由于监控不到位，抛物难以识别等原因，常常无法抓到抛物的不法之徒。所以，高空抛物被称为悬在城市上空的痛。
传统的监控方式是安排专人值守，安排专人值守的方式代价较高，而且大多数高空坠物比较小，人眼可能无法追踪到坠落的物品是从哪里扔出的。
发明内容
基于此，有必要针对上述技术问题，提供一种高空抛物动目标检测方法、检测设备和检测系统，不但能够解决实时监控的人力问题，而且能够精准识别坠物的位置。
第一方面，本申请实施例提供了一种高空抛物动目标检测方法，所述方法包括：
获取若干连续图像；
对若干所述连续图像进行图像预处理；
获取预处理后的图像并进行帧间差分运算，获得差分结果图像；
对所述差分结果图像进行图像处理，获得二值化图像；
对所述二值化图像进行形态学操作，获得形态学结果图像；
对所述形态学结果图像进行连通域分析，确定高空抛物动目标所处位置。
在一些实施例中，所述获取预处理后的图像并进行帧间差分运算，获得差分结果图像包括：
从所述预处理后的图像中获取连续两帧图像，得到前帧图像和后帧图像；
将所述前帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数；
将所述Z张平移图像与所述后帧图像进行差分运算，获得Y张差分图像，其中，Y为整数；
获取所述Y张差分图像相同位置上的像素值的极值作为差分结果；
根据所述差分结果生成差分结果图像。
在一些实施例中，所述获取预处理后的图像并进行帧间差分运算，获得差分结果图像包括：
将所述后帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数；
将所述Z张平移图像与所述前帧图像进行差分运算，获得Y张差分图像，其中，Y为整数；
获取所述Y张差分图像相同位置上的像素值的极值作为差分结果；
根据所述差分结果生成差分结果图像。
在一些实施例中，所述获取所述Y张差分图像相同位置上的像素值的极值作为差分结果，包括：
将所述Y张差分图像相同位置上的像素值进行比较；
获取所述Y张差分图像相同位置上的像素值的极大值作为差分结果；或者，
获取所述Y张差分图像相同位置上的像素值的极小值作为差分结果。
在一些实施例中，所述对所述形态学结果图像进行连通域分析，确定高空抛物动目标所处位置，包括：
对所述形态学结果图像进行连通域分析，获得潜在目标轮廓；
提取所述潜在目标轮廓并进行筛选；
计算筛选后的轮廓的中心坐标；
将所述中心坐标作为高空抛物动目标所处位置。
第二方面，本申请实施例还提供了一种高空抛物动目标检测设备，包括：
至少一个处理器；以及，
与所述至少一个处理器通信连接的存储器；其中，
所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行上述高空抛物动目标检测方法。
第三方面，本申请实施例还提供了一种高空抛物动目标检测系统，包括上述所述的高空抛物动目标检测设备和图像采集设备；
所述图像采集设备与所述高空抛物动目标检测设备连接，所述高空 抛物动目标检测设备用于接收并处理所述图像采集设备发送的图像。
在一些实施例中，所述图像采集设备为红外信号采集设备。
在一些实施例中，所述红外信号采集设备包括星光级相机与红外补光灯的组合，或者红外相机中的至少一种。
第四方面，本申请实施例还提供了一种非易失性计算机可读存储介质，所述计算机可读存储介质存储有计算机可执行指令，当所述计算机可执行指令被处理器所执行时，使所述处理器执行上述高空抛物动目标检测方法。
与现有技术相比，本申请的有益效果是：区别于现有技术的情况，本申请实施例中的高空抛物动目标检测方法、检测设备和检测系统，通过获取若干连续图像，并对所述若干连续图像进行预处理，获得预处理后的图像，接着对预处理后的图像进行帧间差分运算，获得差分结果图像，然后对差分结果图像进行图像处理，得到二值化图像，进一步地，对所述二值化图像进行形态学操作，获得形态学结果图像，最后通过对形态学结果图像进行连通域分析，不但能够识别高空抛物动目标，而且能够确定高空抛物动目标所处位置。
附图说明
一个或多个实施例通过与之对应的附图进行示例性说明，这些示例性说明并不构成对实施例的限定，附图中具有相同参考数字标号的元件表示为类似的元件，除非有特别申明，附图中的图不构成比例限制。
图1是本申请一个实施例中高空抛物动目标检测方法的应用场景示意图；
图2是本申请一个实施例中高空抛物动目标检测方法的流程图；
图3是本申请一个实施例中获得差分结果图像的流程图；
图4是本申请一个实施例中高空抛物动目标检测装置的结构示意图；
图5是本申请一个实施例中高空抛物动目标检测设备的硬件结构示意图。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
需要说明的是，如果不冲突，本申请实施例中的各个特征可以相互结合，均在本申请的保护范围之内。另外，虽然在装置示意图中进行了功能模块划分，在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于装置中的模块划分，或流程图中的顺序执行所示出或描述的步骤。再者，本申请所采用的“第一”、“第二”、“第三”等字样并不对数据和执行次序进行限定，仅是对功能和作用基本相同的相同项或相似项进行区分。
本申请实施例提供的高空抛物动目标检测方法适用于图1所示的应用场景，在本实施例中，所述应用场景为高空抛物动目标检测系统，包括高空抛物动目标检测设备和图像采集设备，所述图像采集设备与所述高空抛物动目标检测设备通信连接，例如可以通过网线连接。可以理解的是，在其他一些实施例中，所述系统还包括交换机。所述图像采集设备还可以通过交换机与所述高空抛物动目标检测设备连接。
图1示例性的示出了图像采集设备20A、图像采集设备20B以及图 像采集设备20C分别通过交换机30与所述高空抛物动目标检测设备10连接。其中，所述图像采集设备20用于将实时采集的图像发送给高空抛物动目标检测设备10进行处理和存储。所述图像采集设备20可以是任意能够检测感应范围内的热红外信号的红外信号采集设备，例如可以为星光级相机与红外补光灯的组合，或者红外相机等。
需要说明的是，本申请实施例提供过的方法还可以进一步的扩展到其他合适的应用环境中，而不限于图1所示的应用环境，在实际的应用过程中，该应用环境还可以包括更多或者更少的图像采集设备、高空抛物动目标检测设备以及交换机。本申请实施例提供的高空抛物动目标检测方法能够监控和识别高空抛物事件的方式与目标位置，并未限定动目标的具体种类，但常见能够识别的有水瓶、烟蒂、其他生活垃圾等。
在一种传统技术中，运动目标检测算法有光流法、背景减除法以及帧间差分法。其中，光流法检测运动目标，其基本思想是赋予图像中的每一个像素点一个速度矢量，从而形成了该图像的运动场。图像上的点和三维物体上的点在某一特定的运动时刻是一一对应的，根据各像素点的速度矢量特征对图像进行动态的分析。若图像中不存在运动目标，那么光流矢量在整个图像区域则是连续变化的，而当物体和图像背景中存在相对运动时，运动物体所形成的速度矢量则必然不同于邻域背景的速度矢量，从而将运动物体的位置检测出来。光流法检测时限不需要知道场景的任何信息就可以准确的计算出运动物体的速度。他不仅能应用于静态背景下的运动目标检测，而且可以用于摄像机运动的情况下实现动态背景运动目标检测。由于光流法要进行迭代运算，精度越高计算量就越大。因此，光流法是一种比较耗时的算法，很难满足实时性的需求。另外，光流法的抗噪性能比较差，当光照发生变化时，即使没有运动发生光流也会存在，容易发生误检。
背景减除法是一种常见的运动对象检测算法，基本思想是利用背景的参数模型来近似背景图像的像素值，将当前帧与背景图像进行差分比较实现对运动区域的检测，其中区别较大的像素区域被认为是运动区域，而区别较小的像素区域被认为是背景区域。背景减除法必须要有背景图像，并且背景图像必须是随着光照或外部环境的变化而实时更新的，因此背景减除法的关键是背景建模及其更新。在天气变化、下雨和反光等条件下，使用背景减除法容易把背景识别成前景。
帧间差分法是最为常用的运动目标检测和分割方法之一，基本原理就是在图像序列相邻两帧或三帧间采用基于像素的时间差分通过闭值化来提取出图像中的运动区域。其背景不积累，且更新速度快、算法简单、计算量小，适合应用于实时性要求较高的场景。算法的不足在于对环境噪声较为敏感，阈值的选择相当关键，选择过低不足以抑制图像中的噪声，过高则忽略了图像中有用的变化。对于比较大的、颜色一致的运动目标，有可能在目标内部产生空洞，无法完整地提取运动目标。
针对上述问题，本申请实施例结合附图提供了如下技术方案。
如图2所示，本申请实施例提供了一种高空抛物动目标检测方法，所述方法由高空抛物动目标检测设备执行，所述方法包括：
步骤202，获取若干连续图像。
在本申请实施例中，将图像采集设备放置于需要进行监控的区域内，24小时对监控区域进行图像采集。高空抛物动目标检测设备实时获取图像采集设备采集的监控图像，并将连续图像转换成视频进行存储，便于后续追踪。其中，获取的若干连续图像的时间间隔为1分钟。可以理解的是，在其他一些实施例中，获取的若干连续图像的时间间隔可根据实际情况自行设置，无需拘泥于本实施例中的限定。
步骤204，对若干所述连续图像进行图像预处理。
在本申请实施例中，图像预处理包括图像颜色空间转换和图像滤波。具体地，获取到的图像为RGB图像，通过特定算法将RGB图像转换为灰度图像，例如通过最大值法、平均值法或者加权平均法将RGB图像转换为灰度图像。为了获得比较干净清晰的图像，进一步地将灰度图像进行滤波处理，由此能够去掉图像中的噪声，有利于后续帧间差分运算、图像处理工作的进行。示例性的，可以将灰度图像进行均值滤波，或者将灰度图像进行中值滤波、高斯滤波或双边滤波等操作，能够有效地去除图像背景中的噪声。
可以理解的是，在本申请其他一些实施例中，还可以将获取到的RGB图像转换为HSV图像，然后对HSV图像进行颜色通道分离，取H通道作为检测通道，从而获得灰度图像，接着按照上述所述方法将灰度图像进行滤波处理，同样可以去除图像背景中的噪声。
步骤206，获取预处理后的图像并进行帧间差分运算，获得差分结果图像。
在本申请实施例中，帧间差分运算为一种改进的图像帧间差分方法，求帧间差分的方式是将图像平移后再作差分计算，多次平移后差分，取差分绝对值的极值作为最终的差分结果。相对比传统的两帧差分法对连续图像取两帧直接作图像差分得到差分图像，以及三帧差分法取连续三帧预处理后的图像，差分前帧图像和中间帧图像得到差分图像1，差分前帧图像和中间帧图像得到差分图像2，差分图像1和图像2得到差分图像3，把差分图像3作为差分结果图像。本申请实施例中的帧间差分方法通过平移再差分的方式，过滤了大部分由于摄像头抖动、下雨、树林和衣物等环境因素产生的干扰，不但改善了传统两帧法和三帧法对环境噪声敏感的缺点，而且能够完整地提取运动目标。
在其他一些实施例中，如图3所示，通过改进的图像帧间差分方法 对预处理后的图像进行图像差分，获得差分结果图像的步骤具体可以包括：
步骤302，从所述预处理后的图像中获取连续两帧图像，得到前帧图像和后帧图像。
在本申请实施例中，前帧图像为前帧灰度图像，后帧图像为后帧灰度图像。具体地，高空抛物动目标检测设备从经过滤波处理的灰度图像中获取连续两帧灰度图像，得到前帧灰度图像和后帧灰度图像。
步骤304，将所述前帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数。
具体地，将前帧灰度图像沿着向量(M,N)平移，得到Z张平移图像，Z张平移图像具体为(2m+1)x(2n+1)。其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m和n为整数，m和n的取值可以是大于等于0且小于图片宽度和高度的整数。示例性的，当m＝10，n＝10，即得到441张平移图像。
可以理解的是，在其他一些实施例中，前帧图像和后帧图像可以互换，平移的对象也可以为后帧图像。具体地，将后帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数。
步骤306，将所述Z张平移图像与所述后帧图像进行差分运算，获得Y张差分图像，其中，Y为整数。
在本申请实施例中，差分图像也成为帧差图像。承接上述例子，将Z张平移图像即441张平移图像与后帧灰度图像进行差分运算即进行帧差，得到441张帧差图像。
在其他一些实施例中，当平移对象为后帧图像时，则将所述Z张平 移图像与所述前帧图像进行差分运算，获得Y张差分图像，其中，Y为整数。
步骤308，获取所述Y张差分图像相同位置上的像素值的极值作为差分结果。
具体地，差分结果是通过获取Y张差分图像相同位置上的像素值的极值得到的。具体地，在一些实施例中，可以通过获取Y张差分图像相同位置上的像素值，然后比较Y张差分图像在每个相同像素位置的值的大小，取极大值作为差分结果。可以理解的是，在其他一些实施例中，还可以取相同位置上的像素值的极小值作为差分结果。
步骤310，根据所述差分结果生成差分结果图像。
具体地，高空抛物动目标检测设备根据差分结果生成差分结果图像，差分结果图像中每个像素的值都是差分图像在相同位置的极值。
当得到差分结果图像后，则执行步骤208，对所述差分结果图像进行图像处理，获得二值化图像。
在本申请实施例中，图像处理包括图像滤波和图像分割。具体地，对差分结果图像做均值滤波，对滤波后的图像进行局部自适应阈值法操作得到二值化图像。可以理解的是，在其他一些实施例中，可以使用高斯滤波、双边滤波、中值滤波中等图像滤波方法的一种或者多种替代本实施例中的均值滤波，可以使用固定阈值法、其他自适应阈值法、区域生长法和分水岭法替代本实施例的局部自适应方法。
步骤210，对所述二值化图像进行形态学操作，获得形态学结果图像。
在本申请实施例中，形态学操作包括腐蚀、膨胀、开运算和闭运算中的一种或者多种。形态学结果图像为膨胀后的二值化图像。具体地，将所述二值化图像进行闭运算操作，获得形态学结果图像即膨胀后的二 值化图像。
步骤212，对所述形态学结果图像进行连通域分析，确定高空抛物动目标所处位置。
具体地，对形态学结果图像即膨胀后的二值化图像进行连通域分析，获得潜在目标轮廓，接着提取所述潜在目标的轮廓，并根据轮廓的面积大小、周长以及圆心率进行筛选，然后通过计算目标轮廓的图像矩得到轮廓的质心，并将所述质心作为轮廓的中心坐标，所述中心坐标即为动目标所在位置，由此能够确定高空抛物动目标所处位置。
在本申请实施例中，通过获取若干连续图像，并对所述若干连续图像进行预处理，获得预处理后的图像，接着对预处理后的图像进行帧间差分运算，获得差分结果图像，然后对差分结果图像进行图像处理，得到二值化图像，进一步地，对所述二值化图像进行形态学操作，获得形态学结果图像，最后通过对形态学结果图像进行连通域分析，不但能够识别高空抛物动目标，而且能够确定高空抛物动目标所处位置。
相应的，本申请实施例还提供了一种高空抛物动目标检测装置400，如图4所示，包括：
获取模块402，用于获取若干连续图像；
第一处理模块404，用于对若干所述连续图像进行图像预处理；
计算模块406，用于获取预处理后的图像并进行帧间差分运算，获得差分结果图像；
第二处理模块408，用于对所述差分结果图像进行图像处理，获得二值化图像；
第三处理模块410，用于对所述二值化图像进行形态学操作，获得形态学结果图像；
分析模块412，用于对所述形态学结果图像进行连通域分析，确定 高空抛物动目标所处位置。
可选的，在装置的其他实施例中，计算模块406具体用于：
从所述预处理后的图像中获取连续两帧图像，得到前帧图像和后帧图像；
将所述前帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数；
将所述Z张平移图像与所述后帧图像进行差分运算，获得Y张差分图像，其中，Y为整数；
获取所述Y张差分图像相同位置上的像素值的极值作为差分结果；
根据所述差分结果生成差分结果图像。
将所述后帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数；
将所述Z张平移图像与所述前帧图像进行差分运算，获得Y张差分图像，其中，Y为整数；
获取所述Y张差分图像相同位置上的像素值的极值作为差分结果；
根据所述差分结果生成差分结果图像。
将所述Y张差分图像相同位置上的像素值进行比较；
获取所述Y张差分图像相同位置上的像素值的极大值作为差分结果；或者，
获取所述Y张差分图像相同位置上的像素值的极小值作为差分结果。
可选的，在装置的其他实施例中，分析模块412具体用于：
对所述形态学结果图像进行连通域分析，获得潜在目标轮廓；
提取所述潜在目标轮廓并进行筛选；
计算筛选后的轮廓的中心坐标；
将所述中心坐标作为高空抛物动目标所处位置。
本申请实施例提供的高空抛物动目标检测装置，通过获取模块获取若干连续图像，并通过第一处理模块对所述若干连续图像进行预处理，获得预处理后的图像，接着通过计算模块对预处理后的图像进行帧间差分运算，获得差分结果图像，然后通过第二处理模块对差分结果图像进行图像处理，得到二值化图像，进一步地，使用第三处理模块对所述二值化图像进行形态学操作，获得形态学结果图像，最后通过分析模块对形态学结果图像进行连通域分析，不但能够识别高空抛物动目标，而且能够确定高空抛物动目标所处位置。
需要说明的上述高空抛物动目标检测装置可执行本申请实施例所提供的高空抛物动目标检测方法，具备执行方法相应的功能模块和有益效果。未在高空抛物动目标检测装置实施例中详尽描述的技术细节，可参考本申请实施例所提供的高空抛物动目标检测方法。
图5是本申请实施例提供的高空抛物动目标检测设备的硬件结构示意图，如图5所示，该高空抛物动目标检测设备500包括：
一个或者多个处理器502以及存储器504，图5中以一个处理器502为例。其中，所述处理器502可以为中央处理器，或者图像处理器，亦或者处理器502为集成中央处理器和图像处理器的功能于一体的处理器。
处理器502和存储器504可以通过总线或者其他方式连接，图5中以通过总线连接为例。
存储器504作为一种非易失性计算机可读存储介质，可用于存储非易失性软件程序、非易失性计算机可执行程序以及模块，如本申请实施 例中的高空抛物动目标检测方法对应的程序/模块(例如，附图4所示的获取模块402、第一处理模块404、计算模块406、第二处理模块408、第三处理模块410以及分析模块412)。处理器502通过运行存储在存储器504中的非易失性软件程序、指令以及模块，从而执行高空抛物动目标检测设备的各种功能应用以及数据处理，即实现上述方法实施例中的高空抛物动目标检测方法。
存储器504可以包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需要的应用程序；存储数据区可存储根据高空抛物动目标检测装置使用所创建的数据等。此外，存储器504可以包括高速随机存取存储器，还可以包括非易失性存储器，例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实施例中，存储器504可选包括相对于处理器502远程设置的存储器，这些远程存储器可以通过网络连接至高空抛物动目标检测装置。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
所述一个或者多个模块存储在所述存储器504中，当被所述一个或者多个高空抛物动目标检测设备执行时，执行上述任意方法实施例中的高空抛物动目标检测方法，例如，执行以上描述的图2中的方法步骤202至步骤212，图3中的方法步骤302至步骤310；实现图4中的模块402至412的功能。
本申请实施例还提供了一种计算机程序产品，包括存储在非易失性计算机可读存储介质上的计算程序，所述计算机程序包括程序指令，当所述程序指令被计算机执行时时，使所述计算机执行：图2中的方法步骤202至步骤212，图3中的方法步骤302至步骤310。
上述产品可执行本申请实施例所提供的方法，具备执行方法相应的 功能模块和有益效果。未在本申请实施例中详尽描述的技术细节，可参考本申请实施例所提供的方法。
本申请实施例中的高空抛物动目标检测设备以多种形式存在，包括但不限于：
(1)移动通信设备:这类设备的特点是具备移动通信功能，并且以提供话音、数据通信为主要目标。这类终端包括:智能手机(例如iPhone)、多媒体手机、功能性手机，以及低端手机等。
(2)超移动个人计算机设备:这类设备属于个人计算机的范畴，有计算和处理功能，一般也具备移动上网特性。这类终端包括:PDA、MI D和UMPC设备等，例如iPad。
(3)便携式娱乐设备:这类设备可以显示和播放多媒体内容。该类设备包括:音频、视频播放器(例如iPod)，掌上游戏机，电子书，以及智能玩具和便携式车载导航设备。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。
通过以上的实施方式的描述，本领域普通技术人员可以清楚地了解到各实施方式可借助程序加通用硬件平台的方式来实现，当然也可以通过硬件。本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程是可以通过计算机程序来指令相关的硬件来完成，所述的程序可存储于一计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory,ROM)或随机存储记忆体(Random Access  Memory,RAM)等。
最后应说明的是：以上实施例仅用以说明本申请的技术方案，而非对其限制；在本申请的思路下，以上实施例或者不同实施例中的技术特征之间也可以进行组合，步骤可以以任意顺序实现，并存在如上所述的本申请的不同方面的许多其它变化，为了简明，它们没有在细节中提供；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的范围。

Claims (10)

  1. 一种高空抛物动目标检测方法，其特征在于，所述方法包括：
    获取若干连续图像；
    对若干所述连续图像进行图像预处理；
    获取预处理后的图像并进行帧间差分运算，获得差分结果图像；
    对所述差分结果图像进行图像处理，获得二值化图像；
    对所述二值化图像进行形态学操作，获得形态学结果图像；
    对所述形态学结果图像进行连通域分析，确定高空抛物动目标所处位置。
  2. 根据权利要求1所述的方法，其特征在于，所述获取预处理后的图像并进行帧间差分运算，获得差分结果图像包括：
    从所述预处理后的图像中获取连续两帧图像，得到前帧图像和后帧图像；
    将所述前帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数；
    将所述Z张平移图像与所述后帧图像进行差分运算，获得Y张差分图像，其中，Y为整数；
    获取所述Y张差分图像相同位置上的像素值的极值作为差分结果；
    根据所述差分结果生成差分结果图像。
  3. 根据权利要求1所述的方法，其特征在于，所述获取预处理后的图像并进行帧间差分运算，获得差分结果图像包括：
    将所述后帧图像沿向量(M,N)平移，得到Z张平移图像，其中，M∈[-m,m],N∈[-n,n],M、N为整数，Z为大于等于1的整数，m、n为整数；
    将所述Z张平移图像与所述前帧图像进行差分运算，获得Y张差分图像，其中，Y为整数；
    获取所述Y张差分图像相同位置上的像素值的极值作为差分结果；
    根据所述差分结果生成差分结果图像。
  4. 根据权利要求2或3所述的方法，其特征在于，所述获取所述Y张差分图像相同位置上的像素值的极值作为差分结果，包括：
    将所述Y张差分图像相同位置上的像素值进行比较；
    获取所述Y张差分图像相同位置上的像素值的极大值作为差分结果；或者，
    获取所述Y张差分图像相同位置上的像素值的极小值作为差分结果。
  5. 根据权利要求4所述的方法，其特征在于，所述对所述形态学结果图像进行连通域分析，确定高空抛物动目标所处位置，包括：
    对所述形态学结果图像进行连通域分析，获得潜在目标轮廓；
    提取所述潜在目标轮廓并进行筛选；
    计算筛选后的轮廓的中心坐标；
    将所述中心坐标作为高空抛物动目标所处位置。
  6. 一种高空抛物动目标检测设备，其特征在于，包括：
    至少一个处理器；以及，
    与所述至少一个处理器通信连接的存储器；其中，
    所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行权利要求1-5任一项所述的方法。
  7. 一种高空抛物动目标检测系统，其特征在于，包括权利要求6所述的高空抛物动目标检测设备和图像采集设备；
    所述图像采集设备与所述高空抛物动目标检测设备连接，所述高空抛物动目标检测设备用于接收并处理所述图像采集设备发送的图像。
  8. 根据权利要求7所述的系统，其特征在于，所述图像采集设备为红外信号采集设备。
  9. 根据权利要求8所述的系统，其特征在于，所述红外信号采集设备包括星光级相机与红外补光灯的组合，或者红外相机中的至少一种。
  10. 一种非易失性计算机可读存储介质，其特征在于，所述计算机可读存储介质存储有计算机可执行指令，当所述计算机可执行指令被处理器所执行时，使所述处理器执行如权利要求1-5任一项所述的方法。
PCT/CN2021/102769 2020-06-28 2021-06-28 一种高空抛物动目标检测方法、检测设备和检测系统 WO2022001961A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010598843.1 2020-06-28
CN202010598843.1A CN111768431B (zh) 2020-06-28 2020-06-28 一种高空抛物动目标检测方法、检测设备和检测系统

Publications (1)

Publication Number Publication Date
WO2022001961A1 true WO2022001961A1 (zh) 2022-01-06

Family

ID=72722192

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/102769 WO2022001961A1 (zh) 2020-06-28 2021-06-28 一种高空抛物动目标检测方法、检测设备和检测系统

Country Status (2)

Country Link
CN (1) CN111768431B (zh)
WO (1) WO2022001961A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114332154A (zh) * 2022-03-04 2022-04-12 英特灵达信息技术(深圳)有限公司 一种高空抛物检测方法及系统
CN115331129A (zh) * 2022-10-14 2022-11-11 彼图科技(青岛)有限公司 一种基于无人机和人工智能的垃圾数据识别方法

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111768431B (zh) * 2020-06-28 2024-04-19 熵康(深圳)科技有限公司 一种高空抛物动目标检测方法、检测设备和检测系统
CN112258573B (zh) * 2020-10-16 2022-11-15 腾讯科技（深圳）有限公司 抛出位置获取方法、装置、计算机设备及存储介质
CN113065454B (zh) * 2021-03-30 2023-01-17 青岛海信智慧生活科技股份有限公司 一种高空抛物目标识别比较的方法及装置
CN113139478A (zh) * 2021-04-27 2021-07-20 广东博智林机器人有限公司 一种高空抛物的检测方法、装置、电子设备及存储介质
CN113382276A (zh) * 2021-06-09 2021-09-10 湖南快乐阳光互动娱乐传媒有限公司 一种图片处理方法及系统
CN114022517A (zh) * 2021-11-29 2022-02-08 北京博思廷科技有限公司 一种基于运动轨迹分析监控视频检测高空抛物的方法
CN114424911B (zh) * 2022-01-23 2024-01-30 深圳银星智能集团股份有限公司 清理方法及移动设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104700430A (zh) * 2014-10-05 2015-06-10 安徽工程大学 机载显示器的运动检测方法
CN107067416A (zh) * 2017-05-11 2017-08-18 南宁市正祥科技有限公司 一种运动目标的检测方法
US9911197B1 (en) * 2013-03-14 2018-03-06 Hrl Laboratories, Llc Moving object spotting by forward-backward motion history accumulation
CN111768431A (zh) * 2020-06-28 2020-10-13 熵康(深圳)科技有限公司 一种高空抛物动目标检测方法、检测设备和检测系统

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7239719B2 (en) * 2003-08-22 2007-07-03 Bbn Technologies Corp. Automatic target detection and motion analysis from image data

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9911197B1 (en) * 2013-03-14 2018-03-06 Hrl Laboratories, Llc Moving object spotting by forward-backward motion history accumulation
CN104700430A (zh) * 2014-10-05 2015-06-10 安徽工程大学 机载显示器的运动检测方法
CN107067416A (zh) * 2017-05-11 2017-08-18 南宁市正祥科技有限公司 一种运动目标的检测方法
CN111768431A (zh) * 2020-06-28 2020-10-13 熵康(深圳)科技有限公司 一种高空抛物动目标检测方法、检测设备和检测系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LI YUAN: "Research on Moving Target Detection and Tracking Algorithm Based on Video", MASTER THESIS, TIANJIN POLYTECHNIC UNIVERSITY, CN, no. 3, 15 March 2020 (2020-03-15), CN , XP055883829, ISSN: 1674-0246 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114332154A (zh) * 2022-03-04 2022-04-12 英特灵达信息技术(深圳)有限公司 一种高空抛物检测方法及系统
CN114332154B (zh) * 2022-03-04 2022-06-14 英特灵达信息技术(深圳)有限公司 一种高空抛物检测方法及系统
CN115331129A (zh) * 2022-10-14 2022-11-11 彼图科技(青岛)有限公司 一种基于无人机和人工智能的垃圾数据识别方法
CN115331129B (zh) * 2022-10-14 2023-03-24 彼图科技(青岛)有限公司 一种基于无人机和人工智能的垃圾数据识别方法

Also Published As

Publication number Publication date
CN111768431B (zh) 2024-04-19
CN111768431A (zh) 2020-10-13

Similar Documents

Publication Publication Date Title
WO2022001961A1 (zh) 一种高空抛物动目标检测方法、检测设备和检测系统
Rakibe et al. Background subtraction algorithm based human motion detection
CN109872341B (zh) 一种基于计算机视觉的高空抛物检测方法与系统
CN109785363A (zh) 一种无人机航拍视频运动小目标实时检测与跟踪方法
CN103093198B (zh) 一种人群密度监测方法及装置
KR101414670B1 (ko) 온라인 랜덤 포레스트 및 파티클 필터를 이용한 열 영상에서의 객체 추적 방법
CN105741319B (zh) 基于盲目更新策略和前景模型的改进视觉背景提取方法
CN111144337B (zh) 火灾检测方法、装置及终端设备
CN104063885A (zh) 一种改进的运动目标检测与跟踪方法
CN102034240A (zh) 一种静态前景检测和跟踪方法
CN114639075B (zh) 一种高空抛物坠物的识别方法、系统及计算机可读介质
Lian et al. A novel method on moving-objects detection based on background subtraction and three frames differencing
CN101715070B (zh) 特定监控视频中的背景自动更新方法
KR101030257B1 (ko) 카메라의 영상을 이용한 보행자 계수 방법 및 장치
CN108765463A (zh) 一种结合区域提取与改进纹理特征的运动目标检测方法
US20130027550A1 (en) Method and device for video surveillance
Zhang et al. An optical flow based moving objects detection algorithm for the UAV
Xu et al. Feature extraction algorithm of basketball trajectory based on the background difference method
Ekinci et al. Background estimation based people detection and tracking for video surveillance
NL2025625B1 (en) System and method for detecting flickering target based on moving camera in specific environment
Karim Construction of a robust background model for moving object detection in video sequence
Cheng et al. A novel improved ViBe algorithm to accelerate the ghost suppression
CN108510527A (zh) 一种基于帧差法和运动点聚类的运动目标检测方法
Kaur Background subtraction in video surveillance
Lande et al. Moving object detection using foreground detection for video surveillance system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21831674

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21831674

Country of ref document: EP

Kind code of ref document: A1