Image sequence optical flow calculation method based on optimized semantic segmentation

A technology of image sequence and semantic segmentation, which is applied in computing, image analysis, image enhancement, etc., and can solve problems such as large errors in optical flow estimation

Active Publication Date: 2020-06-26
NANCHANG HANGKONG UNIVERSITY
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In recent years, with the development of optical flow estimation methods, great progress has been made in optical flow estimation technology for simple scene image seque...

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image sequence optical flow calculation method based on optimized semantic segmentation
  • Image sequence optical flow calculation method based on optimized semantic segmentation
  • Image sequence optical flow calculation method based on optimized semantic segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0055] Below in conjunction with accompanying drawing, the present invention will be further described.

[0056] The technical scheme that present embodiment adopts is: it comprises the following steps: see Figure 6 as shown,

[0057] 1. Input figure 1 and figure 2 are two consecutive frames of pictures in the KITTI image sequence; where figure 1 is the first frame image, figure 2 is the second frame image;

[0058] Use the KITTI image sequence segmented by the deeplabv3+ semantic model to segment the image in two consecutive frames;

[0059] Such as Figure 1-2 As shown, the original image sequence is used to establish a de-mean normalized optical flow model to calculate the foreground and background optical flow between two consecutive frames. The model is as follows:

[0060] E(u,v)=λ·E data (u,v)+E smooth (u,v) (1)

[0061] Data item E in formula (1) data The energy functional is as follows:

[0062]

[0063] where u is the horizontal component of optical...

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

No PUM Login to view more

Abstract

The invention relates to an image sequence optical flow calculation method based on optimized semantic segmentation, which comprises the following steps of: 1, inputting two continuous frames of pictures, and calculating foreground and background optical flows between the two continuous frames by using a mean removal normalization optical flow model; 2, inputting two continuous frames of pictures,and segmenting two continuous frames of label pictures by utilizing a semantic model; 3, taking the optical flow calculated in the step 1 and the label graph in the step 2 as input, and obtaining anoptimized segmentation image by using a full-connection layering algorithm; 4, taking the two continuous frames of pictures in the step 1, the calculated optical flow result and the optimized segmentation result in the step 3 as input quantities, and calculating a final optical flow result by using a foreground optical flow model. According to the method, the side window filtering is added to thefull-connection de-mean normalization model to calculate the optical flow, then the segmentation image prior information optimized by the full-connection hierarchical model is used, and finally the foreground optical flow model is used to optimize the final optical flow, so that the problem of low precision of the calculation result of the sequence optical flow of the image of the illumination mutation scene is solved.

Description

technical field [0001] The invention relates to the technical field of image sequence optical flow calculation, in particular to an image sequence optical flow calculation method based on optimized semantic segmentation. Background technique [0002] Optical flow field is an important method for analyzing moving objects in sequence images. The optical flow field not only contains the motion information of the observed object, but also carries the information of the three-dimensional structure of the optical scene, so the optical flow field plays an important role in different fields: in computer vision, such as target segmentation, recognition, tracking , robot navigation and shape information recovery and other important tasks; optical flow field calculation has important practical significance in industrial and military applications, such as robot vision systems that complete various industrial or military tasks, space satellites based on motion analysis Tracking system; ...

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
no application Login to view more
IPC IPC(8): G06T7/269G06T7/215
CPCG06T7/269G06T7/215G06T2207/10004
Inventor 张聪炫邓士心陈震黎明危水根
Owner NANCHANG HANGKONG UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products