Optical flow calculation method combining image pyramid guidance and cyclic cross attention
A technology of image pyramid and calculation method, applied in the direction of calculation, computer parts, character and pattern recognition, etc., can solve the problem that spatial information cannot be effectively used, the accuracy of motion edge and large displacement optical flow estimation is reduced, and the context of complex motion scenes Insufficient extraction ability and other problems, to achieve the effect of overcoming imbalance, improving accuracy and robustness, and good practicability
Pending Publication Date: 2022-07-29
NANCHANG HANGKONG UNIVERSITY
View PDF0 Cites 4 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] At present, most of the feature extraction methods of optical flow calculation models use feature pyramids, but simply using convolution for feature extraction will make the spatial information in the shallow layer unable to be effectively utilized, resulting in insufficient context extraction capabilities in complex motion scenes, resulting in motion Reduced accuracy of optical flow estimation at edges and large displacements
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
[0026] The technical solutions in the examples of the present invention will be clearly and completely described below with reference to the accompanying drawings in the examples of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
[0027] see Figure 1-Figure 8 , the present invention provides an optical flow calculation method for joint image pyramid subnet guidance and cyclic cross-attention, using the cave_3 sequence image for experimental description:
[0028] 1) Input figure 1 and figure 2 is two consecutive images of the cave_3 image sequence; where: figure 1 is the first frame of image, figure 2 is the second frame image;
[0029] 2) will figure 1 and figure 2 Input to the image pyramid subnet and feature pyramid subnet respectively;
[0030] 3) as image 3 As shown, the image...
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 provides an optical flow calculation method combining image pyramid subnet guidance and cyclic cross attention. The method comprises the following steps of: 1) respectively inputting two continuous frames of images of an image sequence into an image pyramid subnet and a feature pyramid subnet; 2) processing the picture by using the image pyramid subnet; 3) adding and fusing the features extracted by the image pyramid subnet and the features extracted by the same-layer feature pyramid as the input of the next-layer feature pyramid; and 4) inputting the fused features in the last three layers of the feature pyramid into a cyclic cross attention module for context information extraction. According to the optical flow calculation method combining image pyramid subnet guidance and cyclic cross attention, feature information of a moving edge and a large-displacement area of an image sequence is extracted through supplementing of shallow information and accurate extraction of context information, and the accuracy and robustness of optical flow estimation are remarkably improved.
Description
technical field [0001] The invention relates to an optical flow calculation method combining image pyramid guidance and circular cross attention. Background technique [0002] Optical flow is the instantaneous speed of space moving objects moving in the pixel observation plane. It is a method to calculate the motion information of objects between adjacent frames. The direction and speed of motion of image pixels. Recovering the three-dimensional structure and motion of objects from optical flow is one of the most meaningful and challenging tasks faced by existing computer vision research. In computer vision, optical flow plays an important role in target object segmentation, recognition, and tracking. , robot navigation and shape information recovery have very important applications. [0003] At present, most feature extraction methods of optical flow computing models use feature pyramids, but simply using convolution for feature extraction will make the spatial informatio...
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 Applications(China)
IPC IPC(8): G06V10/44G06V10/82G06N3/04
CPCG06V10/44G06V10/82G06N3/045
Inventor 陈震王梓歌张聪炫葛利跃王子旭陈昊黎明胡卫明
Owner NANCHANG HANGKONG UNIVERSITY