Occlusion region detection method based on image sequence optical flow and triangular mesh
A triangular grid and image sequence technology, applied in the field of detection, can solve the problems of robustness and accuracy, which have a great influence on the detection of occlusion areas, and the detection results of occlusion areas do not meet the actual needs, etc., and achieve high accuracy and reliability of occlusion detection , the effect of low complexity
Active Publication Date: 2018-09-21
NANCHANG HANGKONG UNIVERSITY
View PDF2 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] Traditional image sequence occlusion region detection methods usually adopt a strategy based on feature matching. Since the feature matching method only relies on whether a single pixel has a corresponding relationship between two consecutive frames of images to determine the occlusion of a pixel, the robustness of feature point matching And the accuracy has a great influence on the detection of occlusion areas, which often leads to the detection results of occlusion areas not meeting the actual needs
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0028] The present invention will be further described below in conjunction with drawings and embodiments. The occlusion area detection method based on image sequence optical flow and triangular mesh is explained using the RubberWhale image sequence occlusion area detection experiment:
[0029] 1) Figure 1a and Figure 1b It is two consecutive frames of RubberWhale image sequence; where: Figure 1a is the first frame image, Figure 1b is the second frame image;
[0030] figure 2 is the optical flow between two consecutive frames of the calculated RubberWhale image sequence;
[0031] 2) if image 3 As shown, the first frame image is triangulated using the Delaunay triangular mesh;
[0032] 3) Calculate the coordinates of each pixel point in the second frame image according to the result of Delaunay triangular mesh division and the optical flow between images, and establish the corresponding triangular mesh area in the second frame image according to the triangular mesh d...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an occlusion area detection method based on image sequence optical flow and triangular grid. The method calculates the optical flow between two consecutive frames of the image sequence; divides the first frame image into Delaunay triangular grid; calculates the image pixels The brightness change of the point and each triangle area between two frames of images; find the corresponding triangle area of the second frame image according to the triangle of the first frame image, and check whether there are embedded pixels in the corresponding triangle area of the second frame image, and determine that there are occluded triangles area; according to the brightness change of the embedded pixel and the triangular area, it is determined whether the embedded pixel is occluded or the triangular area is occluded; all triangular areas in the image are traversed to detect the occluded area of the entire image. Compared with existing methods, this method has higher occlusion detection accuracy and reliability.
Description
technical field [0001] The invention relates to detection technology, in particular to an occlusion region detection method based on image sequence optical flow and triangular mesh. Background technique [0002] When there is an occlusion phenomenon in the image sequence, the pixels in the occlusion area in the image will appear "occasionally and sometimes absent" in the continuous frame images. This problem will lead to the accuracy of the image sequence moving target detection, tracking and 3D reconstruction results. Reliability faces big challenges. Therefore, image sequence occlusion region detection is an important research content in image processing, computer vision and other research fields, and its research results have important application significance in the fields of robot vision, UAV navigation, and vehicle assisted driving. [0003] Traditional image sequence occlusion region detection methods usually adopt a strategy based on feature matching. Since the feat...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T7/285
Inventor 张聪炫陈震和世熙汪明润江少锋
Owner NANCHANG HANGKONG UNIVERSITY