Moving target detection under static background

Source: Internet
Author: User

Moving object is commonly used in video surveillance field, the aim is to extract the changing region from the background image from the sequence image, and the effective detection of moving region is very important for the post-processing of target classification, tracking and behavior understanding. According to the relationship between the camera and the moving target, it can be divided into the moving object detection under the static background (camera stationary) and the moving object detecting under the dynamic background (the camera also moves simultaneously). In the project, I used the moving object detection in the static background, and I needed to detect the moving objects by fixed camera and complete the gripping action.

Content:

The common methods of moving target detection are generally divided into two categories, one is based on feature, the other is gray-based method. The feature-based method is used to detect moving objects according to the characteristics of images, and it is more useful for large targets and easy to extract features. Gray-based methods are generally used to detect moving objects based on the change of gray level in the image. At present, video-based detection methods mainly based on the inter-frame difference method, based on the optical flow field method, based on the background difference method.

Inter-frame difference method is a detection method based on the strong correlation of two adjacent frames in moving image sequence, and has strong adaptability. However, if the object gray distribution evenly, this method will cause the overlapping parts of the target to form a large hole, serious when the target segmentation is not connected, so that the target is not detected.

The Optical flow field method is a method of detecting segmentation based on the estimation of optical flow, which includes both the motion information of the observed object and the relevant structure information. The discontinuity of the optical flow field can be used to segment images into regions corresponding to different moving objects. But the computation of most optical flow method is complex and time-consuming, which is difficult to meet the demand of real-time monitoring.

Background difference method is the most commonly used method in motion detection, it compares the input image with the background image, and divides the moving target directly according to the change of the statistic information such as gray scale change. The general computational value of the difference method is small and practical, but it is influenced by the external conditions such as light and weather. The basic idea is to subtract the current image from the background, if the pixel difference is greater than a certain threshold, then the pixel is judged to be the point on the moving target. The most important step is the background modeling, which needs to estimate a background model without moving object, and determine the position of the moving target by calculating the difference between the current frame and the background model.

Project, I am going to use the inter-frame difference method to realize moving target detection.

The Inter-frame difference method is used to detect the moving target by making difference or subtraction between the adjacent frames in the sequence image, and using the strong correlation of the adjacent frames in the sequence image to make the change detection. It compares adjacent frames by directly comparing the gray values of pixel points, and then extracts the moving regions in the sequence image by selecting the threshold values. In the sequence image, the change between the K-frame image fk (x, y) and the k+1 frame image fk+1 (x, y) can be expressed as two-value differential image d (x, y), as follows:

  

  , T is a threshold value for differential image thresholding. The part of "1" in the binary image is composed of the part of the two frames corresponding to the gray value of the pixel, which usually includes the moving object and noise, and the portion of "0" is composed of the parts of the corresponding pixel gray values of two frames.

Algorithm flow: Read Video file--image preprocessing (including color image conversion to grayscale image, filter noise reduction-median filter), and moving object detection (morphological filter-expansion, corrosion).

Concrete realization: Adopt C#+EMGU complete. -----------to be continued.

Moving target detection under static background

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.