WO2017028047A1 - 背景模型的提取方法、装置以及图像处理设备 - Google Patents

背景模型的提取方法、装置以及图像处理设备 Download PDF

Info

Publication number
WO2017028047A1
WO2017028047A1 PCT/CN2015/087074 CN2015087074W WO2017028047A1 WO 2017028047 A1 WO2017028047 A1 WO 2017028047A1 CN 2015087074 W CN2015087074 W CN 2015087074W WO 2017028047 A1 WO2017028047 A1 WO 2017028047A1
Authority
WO
WIPO (PCT)
Prior art keywords
pixel
image
background model
pixel value
monitoring area
Prior art date
Application number
PCT/CN2015/087074
Other languages
English (en)
French (fr)
Inventor
杨兵兵
Original Assignee
富士通株式会社
杨兵兵
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 富士通株式会社, 杨兵兵 filed Critical 富士通株式会社
Priority to PCT/CN2015/087074 priority Critical patent/WO2017028047A1/zh
Publication of WO2017028047A1 publication Critical patent/WO2017028047A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/20Analysis of motion
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00Details of television systems
    • H04N5/14Picture signal circuitry for video frequency region

Definitions

  • the present invention relates to the field of graphic image technology, and in particular, to a method and apparatus for extracting a background model and an image processing apparatus.
  • background images are widely used in the field of image monitoring and the like.
  • the difference between the current frame and the reference frame can be compared, thereby detecting a moving object.
  • the reference frame may be referred to as a "background image” or represented using a "background model.”
  • Embodiments of the present invention provide a method, an apparatus, and an image processing apparatus for extracting a background model. It is possible to reduce ghosting during image detection and to detect objects that move at a relatively small speed or are stationary for a period of time.
  • a method for extracting a background model includes:
  • a background model of the monitoring area is obtained according to a frequency of occurrence of a range of pixel values in the multi-frame image.
  • an apparatus for extracting a background model comprising:
  • An image acquisition unit that acquires a multi-frame image of the monitoring area
  • the model obtaining unit obtains a background model of the monitoring area according to a frequency of occurrence of a range of pixel values in the multi-frame image.
  • an image processing apparatus comprising the extraction means of the background model as described above.
  • a computer readable program wherein when the program is executed in an image processing apparatus, the program causes a computer to execute a background model as described above in the image processing apparatus Extraction method.
  • a storage medium storing a computer readable program, wherein the computer readable program causes a computer to perform an extraction method of a background model as described above in an image processing device.
  • An advantageous effect of the embodiment of the present invention is that a background model of the monitoring area is obtained according to the frequency of occurrence of the pixel value range in the multi-frame image.
  • FIG. 1 is a schematic diagram of a method for extracting a background model according to Embodiment 1 of the present invention
  • FIG. 2 is a diagram showing an example of a histogram for counting a certain pixel position according to Embodiment 1 of the present invention
  • FIG. 3 is another schematic diagram of a method for extracting a background model according to Embodiment 1 of the present invention.
  • FIG. 4 is a schematic diagram of an apparatus for extracting a background model according to Embodiment 2 of the present invention.
  • Figure 5 is another schematic diagram of an apparatus for extracting a background model according to Embodiment 2 of the present invention.
  • Fig. 6 is a block diagram showing the configuration of an image processing apparatus according to a third embodiment of the present invention.
  • ghosting may occur in a scene where, for example, when a moving object becomes a stationary object for a period of time (for example, a vehicle waiting for a traffic light), the moving object may be considered to be stationary and updated.
  • a moving object becomes a stationary object for a period of time (for example, a vehicle waiting for a traffic light)
  • the moving object may be considered to be stationary and updated.
  • ghosts will be left when the object moves again.
  • an image monitoring scene in the traffic field will be taken as an example for description.
  • a pixel value for a certain position of the monitoring area (for example, in units of pixels, hereinafter referred to as a pixel position) often appears in this scene (for example, the longest time of occurrence in a period of time), which can be regarded as the background pixel value of the position.
  • the present invention is not limited thereto, and can be applied to other scenarios.
  • Embodiments of the present invention provide a method for extracting a background model.
  • 1 is a schematic diagram of a method for extracting a background model according to an embodiment of the present invention. As shown in FIG. 1, the extraction method includes:
  • Step 101 Acquire a multi-frame image of a monitoring area.
  • Step 102 Obtain a background model of the monitoring area according to a frequency of occurrence of a range of pixel values in the multi-frame image.
  • video information including a plurality of frame images can be obtained by the camera.
  • the camera may be a camera for performing traffic image monitoring, and the monitoring area is continuously captured; however, the present invention is not limited thereto, and may be other image monitoring scenes.
  • steps 101 to 102 can be applied to the initialization process of the background model, and then the background model can be continuously updated.
  • the background model can be performed based on the multi-frame image in the video.
  • an initial background model M 0 can be obtained from 1000 frames of images, and then the background model is continuously updated to obtain background models M 1 , M 2 , . . . , Mi.
  • obtaining the background model of the monitoring area according to the frequency of occurrence of the pixel value range in the multi-frame image may include: counting, for each pixel position of the monitoring area, the pixel location in the Obtaining pixel value range frequency information corresponding to the pixel position in a pixel value in the multi-frame image; and determining a background model of the monitoring area according to the pixel value range frequency information.
  • the range of pixel values should be understood broadly, and the number of pixels in a certain range of pixel values may be plural or one.
  • the range of pixel values may be a case similar to 224-255, or may be a case where there is only a single pixel value such as 255. That is, a single pixel value can be used as a special case of a range of pixel values.
  • 1000 frames of images may be obtained first.
  • the corresponding pixel value in the 0th frame image is 255, and the corresponding pixel value in the first frame image is 80, in the second
  • the corresponding pixel value in the frame image is 255, ..., and the corresponding pixel value in the 999th frame image is 255.
  • the pixel value range frequency information can be obtained by performing statistics on the 1000 pixel values. For example, when the pixel value 255 in the image of a certain frame is obtained, the frequency corresponding to the range of pixel values (for example, 224-255) in which the pixel value 255 is located is incremented by one.
  • the pixel value range frequency information may be represented by a histogram; the abscissa of the histogram is a range of pixel values, and the ordinate of the histogram is a frequency.
  • the present invention is not limited thereto, and may be represented by other forms such as an array.
  • a pixel value range For example, 0-255 can be divided into 8 ranges, and each cylinder corresponds to a range of pixel values in the histogram. Each column has a width of 32 and the column height represents frequency.
  • the present invention is not limited thereto, and a specific form of a range of pixel values may be determined according to actual conditions.
  • FIG. 2 is a diagram showing an example of a histogram for counting a certain pixel position according to an embodiment of the present invention.
  • the abscissa of the histogram is a range of pixel values, and the ordinate of the histogram is frequency.
  • pixel values 81, 79 falling within the pixel value range 64-95
  • 162, 164, and 166 falling in the pixel value range 160
  • the three pixel value ranges (64-95), (160-191), and (224-255) correspond to a frequency of 200 Times, 100 times and 700 times.
  • a plurality of pixel values corresponding to the pixel value range may be averaged to obtain one pixel average value. Therefore, a plurality of pixel average values corresponding to a plurality of pixel value ranges respectively can be obtained, and a background model of the monitoring area is obtained according to the average value.
  • the pixel value range 64-95 has pixel values 81 and 79, then the pixel value range 64-95 can be calculated corresponding to the average value 80; the pixel value range 160-191 has pixels For values 162, 164, and 166, a pixel value range 160-191 corresponding to the average value 164 can be calculated; with pixel values 253 and 255 within the pixel value range 224-255, a pixel value range 224-255 corresponding to the average value 254 can be calculated.
  • the above only schematically illustrates how to calculate the background pixel value.
  • the present invention is not limited thereto, and may be appropriately modified or adjusted. For example, it is possible to calculate only the range of pixel values of the previous one or more frequencies.
  • the plurality of pixel average values are weighted to obtain the background model.
  • the present invention is not limited thereto, and the background model may be constructed using these pixel average values in other ways.
  • the average value of the pixel values with the most occurrences (represented by hiss_max_average) and the average value of the pixel values of the second most frequent occurrence (represented by hiss_secondary_average) may be used as background pixel values. This makes the background model more accurate and more tolerant of noise.
  • one or more frames of the new image of the monitoring area may be acquired to replace one or more old images in the multi-frame image; and the monitoring area is paired according to each new image or old image.
  • the background pixel values are updated.
  • FIG. 3 is another schematic diagram of a method for extracting a background model according to an embodiment of the present invention, schematically showing a case of initializing a background model and updating a background model.
  • the extraction method includes:
  • Step 301 Acquire a multi-frame image of a monitoring area.
  • Step 302 for each pixel position of the monitoring area, counting the pixel position in the multi-frame image a pixel value to obtain a pixel value range frequency information corresponding to the pixel position;
  • Step 303 Determine an initial background model according to the pixel value range frequency information.
  • a queue of length 1000 can be predefined, and 1000 frames of images are placed in the queue. Then, for each pixel position of the monitoring area, when the pixel value of the pixel position in a certain frame image of the 1000 frames is obtained, the frequency corresponding to the pixel value range in which the pixel value is located is incremented by one.
  • multiple frames may be acquired first and then statistics may be performed; or statistics may be performed after each frame is acquired, and then the information of the frame may be pushed into a queue to implement first-in-first. Out.
  • Step 304 Acquire a new image of a frame of the monitoring area
  • the information of the multi-frame image may be organized in the form of a queue, and after obtaining a new image of one frame, the old image of the first obtained frame may be replaced. Therefore, in combination with other factors, fast calculation can be performed to reduce the calculation time; however, the present invention is not limited thereto, and for example, it is also possible to obtain statistics and update after multiple frames.
  • Step 305 Update a background pixel value of the monitoring area according to the new image and the old image.
  • the method may include: for each pixel position of the monitoring area, subtracting a frequency corresponding to a range of pixel values of the pixel position in the old image by 1, and placing the pixel position in the new The frequency corresponding to the range of pixel values in the image is increased by one.
  • the pixel position corresponding to the pixel value in the image of the first frame may be corresponding to the pixel value range.
  • the frequency is decremented by 1, and the frequency corresponding to the pixel value range in which the pixel position in the image of the 1001th frame is located is incremented by one.
  • step 306 it is determined whether to continue the update; if yes, step 304 is performed to retrieve a new image for updating.
  • a frame image update is taken as an example, but the present invention is not limited thereto; for example, statistics and update of pixel value frequency information may be continuously performed, when a preset condition is reached (for example, a statistical N frame image, or After the preset time T) is executed, the background pixel value is updated and the background model is updated.
  • a preset condition for example, a statistical N frame image, or After the preset time T
  • the background pixel value is updated and the background model is updated.
  • the background model of the monitoring area is obtained according to the frequency of occurrence of the pixel value range in the multi-frame image.
  • the embodiment of the present invention provides a device for extracting a background model, and the same content as that of Embodiment 1 will not be described again.
  • the background model extraction apparatus 400 includes:
  • the image obtaining unit 401 acquires a multi-frame image of the monitoring area
  • the model obtaining unit 402 obtains a background model of the monitoring area according to a frequency of occurrence of a range of pixel values in the multi-frame image.
  • the image acquisition unit 401 can obtain a multi-frame image according to the video information obtained by the camera.
  • FIG. 5 is another schematic diagram of an apparatus for extracting a background model according to an embodiment of the present invention.
  • the background model extraction apparatus 500 includes an image acquisition unit 401 and a model obtaining unit 402, as described above.
  • the model obtaining unit 402 may include:
  • a statistic unit 501 for each pixel position of the monitoring area, counting pixel values of the pixel position in the multi-frame image to obtain pixel value range frequency information corresponding to the pixel position;
  • the determining unit 502 determines the background pixel value according to the pixel value range frequency information.
  • the pixel value range frequency information may be represented by a histogram; wherein the abscissa of the histogram is a range of pixel values, and the ordinate of the histogram is frequency, but the invention is not limited thereto.
  • the statistic unit 501 is specifically configured to: when obtaining the pixel value of the pixel position in a certain frame image, add a frequency corresponding to the pixel value range of the pixel value 1.
  • the present invention is not limited thereto, and other statistical methods may be employed, for example.
  • the determining unit 502 is specifically configured to: average a plurality of pixel values corresponding to the plurality of pixel value ranges, and obtain a plurality of pixel average values corresponding to the plurality of pixel value ranges; The pixel average is obtained for the background model.
  • the determining unit 502 may be further configured to: perform weighting processing on the average value of the pixels to obtain the background model.
  • the present invention is not limited thereto, and the background model may be constructed in other ways.
  • the image obtaining unit 401 may be further configured to acquire one or more new frames of the monitoring area to replace one or more old images in the multi-frame image;
  • the extraction device 500 of the background model may further include:
  • the model updating unit 503 updates the background model of the monitoring area according to each frame of new image or old image.
  • the pixel value updating unit 503 may be specifically configured to: for each pixel position of the monitoring area, the frequency corresponding to the pixel value range of the pixel position in each old image of each frame The degree is decremented by 1, and the frequency corresponding to the range of pixel values of the pixel position in the new image of each frame is incremented by one.
  • the present invention is not limited thereto, and other updating methods may be employed.
  • the background model of the monitoring area is obtained according to the frequency of occurrence of the pixel value range in the multi-frame image.
  • An embodiment of the present invention provides an image processing apparatus, where the image processing apparatus includes: a background model extraction apparatus according to Embodiment 2.
  • Fig. 6 is a block diagram showing the configuration of an image processing apparatus according to an embodiment of the present invention.
  • the image processing apparatus 600 may include a central processing unit (CPU) 100 and a memory 110; the memory 110 is coupled to the central processing unit 100.
  • the memory 110 can store various data; in addition, a program for information processing is stored, and the program is executed under the control of the central processing unit 100.
  • the functionality of the background model's extraction device 400 or 500 can be integrated into the central processor 100.
  • the central processing unit 100 may be configured to implement the extraction method of the background model as described in Embodiment 1.
  • the extraction device 400 or 500 of the background model may be configured separately from the central processing unit.
  • the extraction device 400 or 500 of the background model may be configured as a chip connected to the central processing unit 100 through the central processing unit. The functions of the extraction device 400 or 500 that implement the background model are controlled.
  • the image processing apparatus 600 may further include: an input and output unit 120, a display unit 130, and the like; wherein the functions of the above components are similar to those of the prior art, and are not described herein again. It is to be noted that the image processing apparatus 600 does not necessarily have to include all of the components shown in FIG. 6; in addition, the image processing apparatus 600 may further include components not shown in FIG. 6, and reference may be made to the related art.
  • Embodiments of the present invention also provide a computer readable program, wherein when the program is executed in an image processing apparatus, the program causes a computer to perform extraction of a background model as described in Embodiment 1 in the image processing apparatus method.
  • An embodiment of the present invention further provides a storage medium storing a computer readable program, wherein the computer is The reading program causes the computer to execute the extraction method of the background model as described in Embodiment 1 in the image processing apparatus.
  • the above apparatus and method of the present invention may be implemented by hardware or by hardware in combination with software.
  • the present invention relates to a computer readable program that, when executed by a logic component, enables the logic component to implement the apparatus or components described above, or to cause the logic component to implement the various methods described above Or steps.
  • the present invention also relates to a storage medium for storing the above program, such as a hard disk, a magnetic disk, an optical disk, a DVD, a flash memory, or the like.
  • One or more of the functional blocks described in the figures and/or one or more combinations of functional blocks may be implemented as a general purpose processor, digital signal processor (DSP) for performing the functions described herein.
  • DSP digital signal processor
  • ASIC application specific integrated circuit
  • FPGA field programmable gate array
  • One or more of the functional blocks described with respect to the figures and/or one or more combinations of functional blocks may also be implemented as a combination of computing devices, eg, a combination of a DSP and a microprocessor, multiple microprocessors One or more microprocessors in conjunction with DSP communication or any other such configuration.

Landscapes

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

Abstract

一种背景模型的提取方法、装置以及图像处理设备。所述提取方法包括:获取监控区域的多帧图像;根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。由此,可以减小图像检测过程中的鬼影现象,并且能够检测移动速度比较小或者一定时间内静止的物体,图像检测的准确性更高且容忍噪声的能力更强。

Description

背景模型的提取方法、装置以及图像处理设备 技术领域
本发明涉及一种图形图像技术领域,特别涉及一种背景模型的提取方法、装置以及图像处理设备。
背景技术
背景图像的提取被广泛应用在图像监控等领域。例如在检测视频中移动物体时,可以比较当前帧和参考帧的差别,由此检测到运动物体。其中,参考帧可以被称为“背景图像”或者使用“背景模型”进行表示。
目前已经有一些方法来进行背景模型的提取,例如帧差别法(Frame differencing),均值过滤法(Mean filter)以及背景混合模型法(Background Mixture Model)。
应该注意,上面对技术背景的介绍只是为了方便对本发明的技术方案进行清楚、完整的说明,并方便本领域技术人员的理解而阐述的。不能仅仅因为这些方案在本发明的背景技术部分进行了阐述而认为上述技术方案为本领域技术人员所公知。
发明内容
但是,发明人发现,采用目前的背景模型提取方法,经常会出现“鬼影(ghost)”现象;并且一般只能检测到移动速度比较快的运动物体。
本发明实施例提供一种背景模型的提取方法、装置以及图像处理设备。可以减少图像检测过程中的鬼影现象,并且能够检测移动速度比较小或者一段时间内静止的物体。
根据本发明实施例的第一个方面,提供一种背景模型的提取方法,所述提取方法包括:
获取监控区域的多帧图像;
根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。
根据本发明实施例的第二个方面,提供一种背景模型的提取装置,所述提取装置包括:
图像获取单元,获取监控区域的多帧图像;
模型获得单元,根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。
根据本发明实施例的第三个方面,提供一种图像处理设备,所述图像处理设备包括如上所述的背景模型的提取装置。
根据本发明实施例的又一个方面,提供一种计算机可读程序,其中当在图像处理设备中执行所述程序时,所述程序使得计算机在所述图像处理设备中执行如上所述的背景模型的提取方法。
根据本发明实施例的又一个方面,提供一种存储有计算机可读程序的存储介质,其中所述计算机可读程序使得计算机在图像处理设备中执行如上所述的背景模型的提取方法。
本发明实施例的有益效果在于,根据多帧图像中像素值范围出现的频度获得监控区域的背景模型。由此,可以减小图像检测过程中的鬼影现象,并且能够检测移动速度比较小或者一定时间内静止的物体,图像检测的准确性更高且容忍噪声的能力更强。
参照后文的说明和附图,详细公开了本发明的特定实施方式,指明了本发明的原理可以被采用的方式。应该理解,本发明的实施方式在范围上并不因而受到限制。在所附权利要求的条款的范围内,本发明的实施方式包括许多改变、修改和等同。
针对一种实施方式描述和/或示出的特征可以以相同或类似的方式在一个或更多个其它实施方式中使用,与其它实施方式中的特征相组合,或替代其它实施方式中的特征。
应该强调,术语“包括/包含”在本文使用时指特征、整件、步骤或组件的存在,但并不排除一个或更多个其它特征、整件、步骤或组件的存在或附加。
附图说明
参照以下的附图可以更好地理解本发明的很多方面。附图中的部件不是成比例绘制的,而只是为了示出本发明的原理。为了便于示出和描述本发明的一些部分,附图中对应部分可能被放大或缩小。
在本发明的一个附图或一种实施方式中描述的元素和特征可以与一个或更多个其它附图或实施方式中示出的元素和特征相结合。此外,在附图中,类似的标号表示 几个附图中对应的部件,并可用于指示多于一种实施方式中使用的对应部件。
图1是本发明实施例1的背景模型的提取方法的一示意图;
图2是本发明实施例1的对某一像素位置进行统计的直方图的一示例图;
图3是本发明实施例1的背景模型的提取方法的另一示意图;
图4是本发明实施例2的背景模型的提取装置的一示意图;
图5是本发明实施例2的背景模型的提取装置的另一示意图;
图6是本发明实施例3的图像处理设备的一构成示意图。
具体实施方式
参照附图,通过下面的说明书,本发明的前述以及其它特征将变得明显。在说明书和附图中,具体公开了本发明的特定实施方式,其表明了其中可以采用本发明的原则的部分实施方式,应了解的是,本发明不限于所描述的实施方式,相反,本发明包括落入所附权利要求的范围内的全部修改、变型以及等同物。
在图像检测过程中,鬼影现象可能发生在如下场景中:例如,当一个移动物体变成静止物体一段时间(例如等待红绿灯的车辆)之后,该移动物体可能会被认为是静止的而被更新到背景图像或背景模型中,当该物体再次移动时将会留下鬼影。
或者,当一个已经被认为是背景图像区域的物体(例如停车场里已经停了几天的汽车)开始移动时,当该物体开始移动时将会出现鬼影现象。
在本实施例中,将以交通领域的图像监控场景为例进行说明。在该场景下对于监控区域的某一位置(例如以像素为单位,以下称为像素位置)经常出现(例如一段时间中出现时间最长)的像素值,可以认为是该位置的背景像素值。但本发明不限于此,还可以应用到其他的场景中。
实施例1
本发明实施例提供一种背景模型的提取方法。图1是本发明实施例的背景模型的提取方法的一示意图,如图1所示,所述提取方法包括:
步骤101,获取监控区域的多帧图像;
步骤102,根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。
在本实施例中,可以通摄像头获得包括多个帧图像的视频信息。该摄像头可以是用于进行交通图像监控的摄像头,不间断地对监控区域进行拍摄;但本发明不限于此,还可以是其他的图像监控场景。
在本实施例中,步骤101至步骤102可以应用于背景模型的初始化过程,然后该背景模型可以是不断地被更新的。在进行背景模型初始化时,可以根据视频中的多帧图像进行。例如可以根据1000帧图像得到初始背景模型M0,然后不断地对该背景模型进行更新得到背景模型M1,M2,……,Mi。
在本实施例中,根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型,可以包括:对于所述监控区域的每一像素位置,统计所述像素位置在所述多帧图像中的像素值而获得所述像素位置对应的像素值范围频度信息;以及根据所述像素值范围频度信息确定所述监控区域的背景模型。
在本实施例中,像素值范围应被广义地进行理解,处于某一像素值范围的像素值可以是多个也可以是一个。例如像素值范围可以是类似于224-255这样的情况,也可以是例如255这种只有单个像素值的情况。即可以将单个像素值作为像素值范围的一种特例。
例如,可以首先获得1000帧图像,对于某一像素位置[x][y],在第0帧图像中对应的像素值为255,在第1帧图像中对应的像素值为80,在第2帧图像中对应的像素值为255,……,在第999帧图像中对应的像素值为255。
由此,对于监控区域的每一像素位置,可以获得例如1000个像素值,然后可以对于该1000个像素值进行统计而获得像素值范围频度信息。例如,在获得所述像素位置在某一帧图像中的像素值255时,将所述像素值255所在的像素值范围(例如224-255)所对应的频度加1。
在本实施例中,所述像素值范围频度信息可以采用直方图表示;所述直方图的横坐标为像素值范围,所述直方图的纵坐标为频度。但本发明不限于此,例如也可以采用数组等其他形式表示。
以下以像素值范围为例进行说明。例如可以将0-255分为8个范围,在直方图中每个柱体对应一个像素值范围。其中每个柱宽为32,而柱高代表频度。但本发明不限于此,可以根据实际情况确定像素值范围的具体形式。
图2是本发明实施例的对某一像素位置进行统计的直方图的一示例图。如图2所 示,所述直方图的横坐标为像素值范围,所述直方图的纵坐标为频度。对于某一像素位置[x][y],在1000帧图像中出现了例如如下像素值81、79(落在像素值范围64-95内),162、164和166(落在像素值范围160-191内),253和255(落在像素值范围224-255内),三个像素值范围(64-95)、(160-191)以及(224-255)所对应的频度分别为200次、100次和700次。
在本实施例中,对于某一个像素范围,可以将所述像素值范围所对应的多个像素值进行平均而得到一个像素平均值。因此可以获得多个像素值范围分别对应的多个像素平均值,并根据所述平均值获得所述监控区域的背景模型。
例如,对于该像素位置[x][y],像素值范围64-95内具有像素值81和79,则可以计算像素值范围64-95对应平均值80;像素值范围160-191内具有像素值162、164和166,则可以计算像素值范围160-191对应平均值164;像素值范围224-255内具有像素值253和255,则可以计算像素值范围224-255对应平均值254。
值得注意的是,以上仅示意性说明了如何计算背景像素值。但本发明不限于此,还可以进行适当地变型或调整。例如可以仅取前一个或多个频度的像素值范围进行计算。
在本实施例中,将所述多个像素平均值进行加权处理以获得所述背景模型。但本发明不限于此,还可以使用其他的方式利用这些像素平均值构建背景模型。
例如,可以将出现次数最多的像素值的平均值(用hists_max_average表示)以及出现次数第二多的像素值的平均值(用hists_secondary_average表示)均作为背景像素值。由此使得背景模型的准确性更高且容忍噪声的能力更强。
以上对背景模型的生成进行了示意性说明,但本发明不限于此。
在本实施例中,可以获取所述监控区域一帧或多帧新图像,来替换所述多帧图像中的一帧或多帧旧图像;根据每帧新图像或旧图像对所述监控区域的背景像素值进行更新。
图3是本发明实施例的背景模型的提取方法的另一示意图,示意性示出了初始化背景模型和更新背景模型的情况。
如图3所示,所述提取方法包括:
步骤301,获取监控区域的多帧图像;
步骤302,对于监控区域的每一像素位置,统计所述像素位置在所述多帧图像中 的像素值而获得所述像素位置对应的像素值范围频度信息;
步骤303,根据所述像素值范围频度信息确定初始背景模型。
具体如何构建背景模型的方法如上所述。
例如,可以预先定义一个长度为1000的队列,将1000帧图像放入该队列中。然后对于监控区域的每一个像素位置,在获得所述像素位置在该1000帧中的某一帧图像中的像素值时,将所述像素值所在的像素值范围所对应的频度加1。
值得注意的是,在以上的步骤中,可以是首先获取多个帧然后进行统计;也可以是每获取一帧后即进行统计,然后可以将该帧的信息压入一个队列中以实现先入先出。
步骤304,获取监控区域的一帧新图像;
在本实施例中,多帧图像的信息可以以队列的形式组织,在获得一帧新图像后,可以将最先获得的一帧旧图像替换掉。由此结合其他因素可以进行快速计算,减少计算时间;但本发明不限于此,例如也可以获得多帧之后再进行统计并更新。
步骤305,根据新图像和旧图像对所述监控区域的背景像素值进行更新;
其中,具体可以包括:对于所述监控区域的每一像素位置,将所述像素位置在所述旧图像中的像素值范围所对应的频度减1,以及将所述像素位置在所述新图像中的像素值范围所对应的频度加1。
例如,在取得一帧新的图像(即第1001帧)之后,对于所述监控区域的某一像素位置,可以将该像素位置在第1帧图像中的像素值所在的像素值范围所对应的频度减1,将该像素位置在第1001帧图像中的像素值所在的像素值范围所对应的频度加1。
步骤306,确定是否继续进行更新;如果是则执行步骤304,重新获得新的图像进行更新。
在图3中以一帧图像更新为例进行了说明,但本发明不限于此;例如还可以不断地进行统计并更新像素值频度信息,在达到预设条件(例如统计N帧图像,或者执行预设时间T)之后再更新背景像素值并进行背景模型的更新。
由上述实施例可知,根据多帧图像中像素值范围出现的频度获得监控区域的背景模型。由此,可以减小图像检测过程中的鬼影现象,并且能够检测移动速度比较小或者一定时间内静止的物体,图像检测的准确性更高且容忍噪声的能力更强。
实施例2
本发明实施例提供一种背景模型的提取装置,与实施例1相同的内容不再赘述。
图4是本发明实施例的背景模型的提取装置的一示意图,如图4所示,所述背景模型的提取装置400包括:
图像获取单元401,获取监控区域的多帧图像;
模型获得单元402,根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。
在本实施例中,图像获取单元401可根据摄像头获得的视频信息获得多帧图像。
图5是本发明实施例的背景模型的提取装置的另一示意图,如图5所示,所述背景模型的提取装置500包括:图像获取单元401、模型获得单元402,如上所述。
如图5所示,所述模型获得单元402可以包括:
统计单元501,对于所述监控区域的每一像素位置,统计所述像素位置在所述多帧图像中的像素值而获得所述像素位置对应的像素值范围频度信息;
确定单元502,根据所述像素值范围频度信息确定所述背景像素值。
其中,所述像素值范围频度信息可以采用直方图表示;其中所述直方图的横坐标为像素值范围,所述直方图的纵坐标为频度,但本发明不限于此。
在本实施例中,所述统计单元501具体可以用于:在获得所述像素位置在某一帧图像中的像素值时,将所述所述像素值所在像素值范围所对应的频度加1。但本发明不限于此,例如还可以采用其他的统计方式。
在本实施例中,所述确定单元502具体可以用于:将多个所述像素值范围所对应的像素值进行平均,获得多个所述像素值范围对应的多个像素平均值;根据得到的所述像素平均值获得所述背景模型。
此外,所述确定单元502具体还可以用于:将所述像素平均值进行加权处理以获得所述背景模型。但本发明不限于此,还可以采用其他的方式进行背景模型的构建。
在本实施例中,所述图像获取单元401还可以用于获取所述监控区域一帧或多帧新图像以替换所述多帧图像中的一帧或多帧旧图像;
如图5所示,所述背景模型的提取装置500还可以包括:
模型更新单元503,根据每帧新图像或旧图像对所述监控区域的背景模型进行更新。
在本实施例中,所述像素值更新单元503具体可以用于:对于所述监控区域的每一像素位置,将所述像素位置在所述每帧旧图像中的像素值范围所对应的频度减1,以及将所述像素位置在所述每帧新图像中的像素值范围所对应的频度加1。但本发明不限于此,还可以采用其他的更新方式。
由上述实施例可知,根据多帧图像中像素值范围出现的频度获得监控区域的背景模型。由此,可以减小图像检测过程中的鬼影现象,并且能够检测移动速度比较小或者一定时间内静止的物体,图像检测的准确性更高且容忍噪声的能力更强。
实施例3
本发明实施例提供一种图像处理设备,所述图像处理设备包括:如实施例2所述的背景模型的提取装置。
图6是本发明实施例的图像处理设备的一构成示意图。如图6所示,图像处理设备600可以包括:中央处理器(CPU)100和存储器110;存储器110耦合到中央处理器100。其中该存储器110可存储各种数据;此外还存储信息处理的程序,并且在中央处理器100的控制下执行该程序。
在一个实施方式中,背景模型的提取装置400或500的功能可以被集成到中央处理器100中。其中,中央处理器100可以被配置为实现如实施例1所述的背景模型的提取方法。
在另一个实施方式中,背景模型的提取装置400或500可以与中央处理器分开配置,例如可以将背景模型的提取装置400或500配置为与中央处理器100连接的芯片,通过中央处理器的控制来实现背景模型的提取装置400或500的功能。
此外,如图6所示,图像处理设备600还可以包括:输入输出单元120和显示单元130等;其中,上述部件的功能与现有技术类似,此处不再赘述。值得注意的是,图像处理设备600也并不是必须要包括图6中所示的所有部件;此外,图像处理设备600还可以包括图6中没有示出的部件,可以参考现有技术。
本发明实施例还提供一种计算机可读程序,其中当在图像处理设备中执行所述程序时,所述程序使得计算机在所述图像处理设备中执行如实施例1所述的背景模型的提取方法。
本发明实施例还提供一种存储有计算机可读程序的存储介质,其中所述计算机可 读程序使得计算机在图像处理设备中执行如实施例1所述的背景模型的提取方法。
本发明以上的装置和方法可以由硬件实现,也可以由硬件结合软件实现。本发明涉及这样的计算机可读程序,当该程序被逻辑部件所执行时,能够使该逻辑部件实现上文所述的装置或构成部件,或使该逻辑部件实现上文所述的各种方法或步骤。本发明还涉及用于存储以上程序的存储介质,如硬盘、磁盘、光盘、DVD、flash存储器等。
针对附图中描述的功能方框中的一个或多个和/或功能方框的一个或多个组合,可以实现为用于执行本申请所描述功能的通用处理器、数字信号处理器(DSP)、专用集成电路(ASIC)、现场可编程门阵列(FPGA)或者其它可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件或者其任意适当组合。针对附图描述的功能方框中的一个或多个和/或功能方框的一个或多个组合,还可以实现为计算设备的组合,例如,DSP和微处理器的组合、多个微处理器、与DSP通信结合的一个或多个微处理器或者任何其它这种配置。
以上结合具体的实施方式对本发明进行了描述,但本领域技术人员应该清楚,这些描述都是示例性的,并不是对本发明保护范围的限制。本领域技术人员可以根据本发明的精神和原理对本发明做出各种变型和修改,这些变型和修改也在本发明的范围内。

Claims (17)

  1. 一种背景模型的提取方法,所述提取方法包括:
    获取监控区域的多帧图像;
    根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。
  2. 根据权利要求1所述的提取方法,其中,根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型包括:
    对于所述监控区域的每一像素位置,统计所述像素位置在所述多帧图像中的像素值而获得所述像素位置对应的像素值范围频度信息;
    根据所述像素值范围频度信息确定所述背景模型。
  3. 根据权利要求2所述的提取方法,其中,所述像素值范围频度信息采用直方图表示;其中所述直方图的横坐标为像素值范围,所述直方图的纵坐标为频度。
  4. 根据权利要求3所述的提取方法,其中,统计所述像素位置在所述多帧图像中的像素值包括:在获得所述像素位置在某一帧图像中的像素值时,将所述像素值所在的像素值范围所对应的频度加1。
  5. 根据权利要求2所述的提取方法,其中,根据所述像素值范围频度信息确定所述背景模型包括:
    将多个所述像素值范围所对应的像素值进行平均,获得多个所述像素值范围对应的多个像素平均值;
    根据得到的所述像素平均值获得所述背景模型。
  6. 根据权利要求5所述的提取方法,其中,所述提取方法还包括:
    将所述多个像素平均值进行加权处理以获得所述背景模型。
  7. 根据权利要求1所述的提取方法,其中,所述提取方法还包括:
    获取所述监控区域一帧或多帧新图像以替换所述多帧图像中的一帧或多帧旧图像;以及
    根据每帧新图像或旧图像对所述监控区域的背景模型进行更新。
  8. 根据权利要7所述的提取方法,其中,根据每帧新图像或旧图像对所述监控区域的背景像素值进行更新包括:
    对于所述监控区域的每一像素位置,将所述像素位置在所述每帧旧图像中的像素 值范围所对应的频度减1,以及将所述像素位置在所述每帧新图像中的像素值范围所对应的频度加1。
  9. 一种背景模型的提取装置,所述提取装置包括:
    图像获取单元,获取监控区域的多帧图像;
    模型获得单元,根据所述多帧图像中像素值范围出现的频度获得所述监控区域的背景模型。
  10. 根据权利要求9所述的提取装置,其中,所述模型获得单元包括:
    统计单元,对于所述监控区域的每一像素位置,统计所述像素位置在所述多帧图像中的像素值而获得所述像素位置对应的像素值范围频度信息;
    确定单元,根据所述像素值范围频度信息确定所述背景像素值。
  11. 根据权利要求10所述的提取装置,其中,所述像素值范围频度信息采用直方图表示;其中所述直方图的横坐标为像素值范围,所述直方图的纵坐标为频度。
  12. 根据权利要11所述的提取装置,其中,所述统计单元用于:在获得所述像素位置在某一帧图像中的像素值时,将所述像素值所在的像素值范围所对应的频度加1。
  13. 根据权利要求10所述的提取装置,其中,所述确定单元用于:将多个所述像素值范围所对应的像素值进行平均,获得多个所述像素值范围对应的多个像素平均值;根据得到的所述像素平均值获得所述背景模型。
  14. 根据权利要求13所述的提取装置,其中,所述确定单元还用于:将所述像素平均值进行加权处理以获得所述背景模型。
  15. 根据权利要求9所述的提取装置,其中,所述图像获取单元还用于获取所述监控区域一帧或多帧新图像以替换所述多帧图像中的一帧或多帧旧图像;
    所述提取装置还包括:
    模型更新单元,根据每帧新图像或旧图像对所述监控区域的背景模型进行更新。
  16. 根据权利要求15所述的提取装置,其中,所述模型更新单元用于:
    对于所述监控区域的每一像素位置,将所述像素位置在所述每帧旧图像中的像素值范围所对应的频度减1,以及将所述像素位置在所述每帧新图像中的像素值范围所对应的频度加1。
  17. 一种图像处理设备,包括如权利要求9所述的背景模型的提取装置。
PCT/CN2015/087074 2015-08-14 2015-08-14 背景模型的提取方法、装置以及图像处理设备 WO2017028047A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2015/087074 WO2017028047A1 (zh) 2015-08-14 2015-08-14 背景模型的提取方法、装置以及图像处理设备

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2015/087074 WO2017028047A1 (zh) 2015-08-14 2015-08-14 背景模型的提取方法、装置以及图像处理设备

Publications (1)

Publication Number Publication Date
WO2017028047A1 true WO2017028047A1 (zh) 2017-02-23

Family

ID=58050443

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/087074 WO2017028047A1 (zh) 2015-08-14 2015-08-14 背景模型的提取方法、装置以及图像处理设备

Country Status (1)

Country Link
WO (1) WO2017028047A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110033425A (zh) * 2018-01-10 2019-07-19 富士通株式会社 干扰区域检测装置及方法、电子设备
CN111080583A (zh) * 2019-12-03 2020-04-28 上海联影智能医疗科技有限公司 医学图像检测方法、计算机设备和可读存储介质
CN111179299A (zh) * 2018-11-09 2020-05-19 珠海格力电器股份有限公司 一种图像处理方法及装置
CN112084880A (zh) * 2020-08-14 2020-12-15 江铃汽车股份有限公司 一种图像处理方法、装置、存储介质及设备
CN113310987A (zh) * 2020-02-26 2021-08-27 保定市天河电子技术有限公司 一种隧道衬砌表面检测***及方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000251079A (ja) * 1999-03-03 2000-09-14 Nippon Hoso Kyokai <Nhk> 動画像オブジェクト抽出装置
CN101068342A (zh) * 2007-06-05 2007-11-07 西安理工大学 基于双摄像头联动结构的视频运动目标特写跟踪监视方法
CN101854467A (zh) * 2010-05-24 2010-10-06 北京航空航天大学 一种视频分割中阴影的自适应检测及消除方法
CN102663746A (zh) * 2012-03-23 2012-09-12 长安大学 一种基于视频的背景检测方法
CN103136537A (zh) * 2012-12-12 2013-06-05 惠州学院 一种基于支持向量机的车型识别方法
CN103312960A (zh) * 2012-03-09 2013-09-18 欧姆龙株式会社 图像处理装置、图像处理方法
CN103714703A (zh) * 2013-12-17 2014-04-09 重庆凯泽科技有限公司 一种基于视频图像处理的车流检测算法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000251079A (ja) * 1999-03-03 2000-09-14 Nippon Hoso Kyokai <Nhk> 動画像オブジェクト抽出装置
CN101068342A (zh) * 2007-06-05 2007-11-07 西安理工大学 基于双摄像头联动结构的视频运动目标特写跟踪监视方法
CN101854467A (zh) * 2010-05-24 2010-10-06 北京航空航天大学 一种视频分割中阴影的自适应检测及消除方法
CN103312960A (zh) * 2012-03-09 2013-09-18 欧姆龙株式会社 图像处理装置、图像处理方法
CN102663746A (zh) * 2012-03-23 2012-09-12 长安大学 一种基于视频的背景检测方法
CN103136537A (zh) * 2012-12-12 2013-06-05 惠州学院 一种基于支持向量机的车型识别方法
CN103714703A (zh) * 2013-12-17 2014-04-09 重庆凯泽科技有限公司 一种基于视频图像处理的车流检测算法

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110033425A (zh) * 2018-01-10 2019-07-19 富士通株式会社 干扰区域检测装置及方法、电子设备
CN110033425B (zh) * 2018-01-10 2023-03-28 富士通株式会社 干扰区域检测装置及方法、电子设备
CN111179299A (zh) * 2018-11-09 2020-05-19 珠海格力电器股份有限公司 一种图像处理方法及装置
CN111080583A (zh) * 2019-12-03 2020-04-28 上海联影智能医疗科技有限公司 医学图像检测方法、计算机设备和可读存储介质
CN111080583B (zh) * 2019-12-03 2024-02-27 上海联影智能医疗科技有限公司 医学图像检测方法、计算机设备和可读存储介质
CN113310987A (zh) * 2020-02-26 2021-08-27 保定市天河电子技术有限公司 一种隧道衬砌表面检测***及方法
CN112084880A (zh) * 2020-08-14 2020-12-15 江铃汽车股份有限公司 一种图像处理方法、装置、存储介质及设备

Similar Documents

Publication Publication Date Title
WO2017028047A1 (zh) 背景模型的提取方法、装置以及图像处理设备
WO2017084094A1 (zh) 烟雾检测装置、方法以及图像处理设备
JP6309111B2 (ja) 撮像条件を検出するための方法および装置
CN109791695B (zh) 基于图像块的运动向量确定所述块的方差
US9536321B2 (en) Apparatus and method for foreground object segmentation
WO2017215527A1 (zh) Hdr场景侦测方法、装置和计算机存储介质
CN107909569B (zh) 一种花屏检测方法、花屏检测装置及电子设备
WO2018228310A1 (zh) 图像处理方法、装置及终端
US10666874B2 (en) Reducing or eliminating artifacts in high dynamic range (HDR) imaging
US9904853B2 (en) Monitoring camera device and related region-based motion detection method
US10540546B2 (en) Image processing apparatus, control method, and storage medium
WO2018058530A1 (zh) 目标检测方法、装置以及图像处理设备
US11074742B2 (en) Image processing apparatus, image processing method, and storage medium
WO2018068300A1 (zh) 图像处理方法和装置
US10593044B2 (en) Information processing apparatus, information processing method, and storage medium
CN113657434A (zh) 人脸人体关联方法、***以及计算机可读存储介质
WO2019120025A1 (zh) 照片的调整方法、装置、存储介质及电子设备
CN110689496B (zh) 降噪模型的确定方法、装置、电子设备和计算机存储介质
CN110114801B (zh) 图像前景检测装置及方法、电子设备
JP7163718B2 (ja) 干渉領域検出装置と方法及び電子機器
WO2017028029A1 (zh) 背景模型的提取方法、装置以及图像处理设备
US11373277B2 (en) Motion detection method and image processing device for motion detection
CN112084880A (zh) 一种图像处理方法、装置、存储介质及设备
JP7251425B2 (ja) 付着物検出装置および付着物検出方法
CN109558881B (zh) 一种基于计算机视觉的危岩崩塌监控方法

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: 15901252

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: 15901252

Country of ref document: EP

Kind code of ref document: A1