Unlock instant, AI-driven research and patent intelligence for your innovation.

A Continuous Action Recognition Method Based on Improved Viterbi Algorithm

A technology of Viterbi algorithm and action recognition, applied in the field of video processing

Inactive Publication Date: 2017-10-17
BEIJING INSTITUTE OF TECHNOLOGYGY +1
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In this way, from the perspective of the observer, only the observation value can be seen, unlike the one-to-one correspondence between the observation value and the state in the Markov chain model

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
  • A Continuous Action Recognition Method Based on Improved Viterbi Algorithm
  • A Continuous Action Recognition Method Based on Improved Viterbi Algorithm
  • A Continuous Action Recognition Method Based on Improved Viterbi Algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0094] The implementation of the method of the present invention will be described in detail below in conjunction with the accompanying drawings and specific examples.

[0095] The training data includes 400 whole-body motion videos of the human body. The whole-body motion videos of the human body are RGB-D format videos captured by Microsoft's KINECT tool. There are 8 actions in total, and 50 videos are collected for each action. These full-body action videos only contain a single action, and the videos of the same action are taken by performers with different body shapes. There are 10 test data, and each test data contains more than 2 continuous actions. Continuous actions are randomly combined from 10 actions in the training data.

[0096] In this embodiment, the continuous action recognition method based on the improved Viterbi algorithm is used to perform action recognition on the video in the test data, and the operation steps include a training process and a recognitio...

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 belongs to the field of video processing, and in particular relates to a continuous motion recognition method based on an improved Viterbi algorithm, which is used for efficiently recognizing continuous motions of the whole body of a human body. Firstly, the movement difference caused by factors such as human rotation and translation is eliminated by establishing a coordinate system based on human joints. Secondly, the coordinates of the human skeleton are further processed according to the constraints of the human body model to eliminate the difference due to the human body shape, and then the data of the joint points of the human body are encoded by K-means, and then the corresponding hidden Markov model is trained for each action. Then, the continuous action sequence is segmented through the change of active joints, speed, and angle, and then better candidate actions are screened out for each subsequence. Finally, the best path is found through the improved Viterbi algorithm, and then the best action sequence is obtained. Action sequence. Compared with existing methods, this method can efficiently identify actions contained in continuous action sequences.

Description

technical field [0001] The invention belongs to the field of video processing, and in particular relates to a continuous motion recognition method based on an improved Viterbi algorithm, which is used for efficiently recognizing continuous motions of the whole body of a human body. Background technique [0002] With the rapid development of computer applications, human behavior recognition based on video sequences has gradually become a research hotspot in the field of computer vision, and has been widely used in human-machine interface systems, smart home applications, intelligent monitoring, and motion analysis. [0003] Scholars at home and abroad have done a lot of basic research on the problem of action recognition, which has high application value. Commonly used action recognition methods are: HMM-based methods, DTW-based methods, SVM-based methods, and neural network-based methods. For single action recognition, these methods have achieved relatively satisfactory res...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06V40/23G06V30/194
Inventor 张磊白栋天黄华
Owner BEIJING INSTITUTE OF TECHNOLOGYGY