Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A human detection method based on sample transfer learning

A transfer learning and human detection technology, applied in the field of human detection based on sample transfer learning

Active Publication Date: 2016-09-21
INST OF AUTOMATION CHINESE ACAD OF SCI
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to solve the problem of human body detection in complex scenes. To this end, the present invention provides a human body detection method based on sample transfer learning

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 human detection method based on sample transfer learning
  • A human detection method based on sample transfer learning
  • A human detection method based on sample transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0014] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be described in further detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0015] figure 1 It is a flow chart of the human body detection method based on example transfer learning proposed by the present invention, such as figure 1 As shown, the human detection method based on example transfer learning includes the following steps:

[0016] Step S1, normalize each positive sample in the training set, and extract gradient histogram features from the obtained normalized images; for the negative samples in the training set, randomly select from pictures that do not contain human bodies An image of the same size as the normalized image, and extract its gradient histogram features, so as to obtain the multi-dimensional human body features of the positive samples and negative samples respectively;

...

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 discloses a human body detection method based on sample transfer learning. The human body detection method based on sample transfer learning comprises the following steps: multi-dimensional human body features of positive samples and negative samples in a training set are extracted; a support vector machine is used for training to obtain an initial human body detector; the human body detector is divided into a plurality of small grids, wherein each small grid represents a small component, and the weight of the corresponding grid is used for representing the small component; the weight of the grids is updated; human body detection is conducted on an image to be detected in a scanning mode through a multi-scale sliding frame, and therefore a human body detection result is obtained. According to the human body detection method based on sample transfer learning, the detector can be structurally adjusted in a self-adaptive mode so as to be matched with the structure of the samples in a detection frame and reckon in matching loss into a scoring function, and therefore the human body detection method can be used for conducting structural transferring on each sample in a self-adaptive mode and processing shape change in a human body and other conditions.

Description

technical field [0001] The invention belongs to the technical field of intelligent video monitoring, and in particular relates to a human body detection method based on sample transfer learning. Background technique [0002] Human detection is the basis of high-level semantic understanding in video surveillance, and the advancement of this technology can greatly promote the development of video surveillance technology. Although pedestrian detection technology in specific scenes has been well developed, there are still many challenges in human detection technology in generalized scenes, mainly due to the change of human appearance and posture. In order to solve the changes in the appearance of the human body, many researchers have proposed features that are robust to background, lighting, and clothing colors. Dalal et al. propose a gradient histogram feature that is robust to illumination. Tuzel et al. utilize magnitude, gradient, and spatial location information to further...

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): G06K9/00G06K9/66
Inventor 王春恒周文肖柏华张重
Owner INST OF AUTOMATION CHINESE ACAD OF SCI
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com