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

Image characteristic extraction method, pedestrian detection method and device

An image feature extraction and pedestrian detection technology, applied in the field of image processing, can solve problems such as limited identification ability, and achieve the effect of rich information, high accuracy, and enhanced identification.

Inactive Publication Date: 2015-12-23
PEKING UNIV
View PDF7 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Therefore, the technical problem to be solved by the present invention is to overcome the defects in the prior art that image features are based on low-level image pixel extraction and limited discrimination ability, thereby providing an image feature extraction method and a pedestrian detection method

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 characteristic extraction method, pedestrian detection method and device
  • Image characteristic extraction method, pedestrian detection method and device
  • Image characteristic extraction method, pedestrian detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0052] This embodiment provides an image feature extraction method, such as figure 1 shown, including the following steps:

[0053] S1. Acquire mid-image features in the target image.

[0054] Among them, the middle-level features refer to the image histogram features constructed based on visual keywords and bag-of-words models. Generally, the underlying local features (such as SIFT) are extracted first, and then clustered to obtain a visual keyword dictionary. Finally, based on these visual Keywords quantify the original underlying local features and obtain them in the form of bag-of-words histograms. Compared with commonly used image bottom-level features, also known as basic features (such as color, gradient), middle-level features can represent richer patterns and information, and also provide a basis for obtaining higher-level semantic information.

[0055] S2. Obtain initial values ​​of hidden semantic features.

[0056] In this embodiment, the hidden semantic feature...

Embodiment 2

[0080] This embodiment provides a specific implementation manner of an image feature extraction method, including the following process.

[0081] In the first step, basic features are extracted for each sample.

[0082] For each sample, local feature descriptors are first extracted at fixed step intervals, and then the k-means clustering algorithm is used to quantify these feature descriptors into a fixed-dimensional histogram based on the bag-of-words model as the basic feature of the sample. Specifically, it can be described as: given a dictionary of visual features Where M is the dimension of the dictionary, and w is the visual keyword in the dictionary, then each sample can be expressed as a histogram based on the visual word bag model {freq(w i ): i=1, 2, ..., M}, where freq(w i ) is the visual keyword w i frequency in each sample.

[0083] The second step is to extract hidden semantic features.

[0084] The extraction of hidden semantic features is a specific spars...

Embodiment 3

[0108] This embodiment provides a method for pedestrian detection, applying the image feature extraction method in Embodiment 1 to pedestrian detection, such as figure 2 shown, including the following steps:

[0109] S11. Extract basic features for each training sample. It is the same as that in Embodiment 2 and will not be repeated here.

[0110] S12. Obtain the implicit semantic feature of the basic feature, the extraction method is the same as that in Embodiments 1 and 2, which will not be repeated here.

[0111] S13. Establish a pedestrian detection model according to the hidden semantic features.

[0112] After obtaining the hidden semantic features X of the training samples, in this scheme, these hidden semantic features are counted in the form of histograms in the regular image units in a manner similar to the HOG feature. Specifically, for each non-zero x i , using the bilinear interpolation method to convert its absolute value|x i |Assigned to one of the four ce...

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 provides an image characteristic extraction method and a pedestrian detection method. The image characteristic extraction method comprises steps that, middle-level image characteristics of a target image are acquired, an initial value of a hidden meaning characteristic is then acquired, a reconstruction error constraint term, a sparsity constraint term and a discriminability constraint term are respectively determined according to the middle-level image characteristics and the initial value of the hidden meaning characteristic; and the hidden meaning characteristic is determined according to the reconstruction error constraint term, the sparsity constraint term and the discriminability constraint term. According to the image characteristic extraction method, higher-level meaning information and more powerful discriminability information are both considered as crucial factors for determining characteristic performance, a characteristic containing richer information and more powerful discriminability is acquired through respectively optimizing a proposed hidden meaning characteristic learning problem and the largest discriminability constraint, and thereby higher pedestrian detection accuracy is realized.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an image feature extraction method and a pedestrian detection method. Background technique [0002] The application of image information is increasingly widespread, and the extraction of image features has become one of the key technologies. Whether the image features are properly expressed has an important impact on the results of image detection and recognition. For example, accurate and reliable detection of pedestrians is an important link in many pedestrian-based computer vision and pattern recognition applications, such as video surveillance, assisted driving, autonomous robot navigation, and more. Thanks to powerful feature representations, robust pedestrian modeling methods, and effective detection strategies, current pedestrian detection techniques have achieved significant progress in both accuracy and speed, and their performance has improved over the past decades. by...

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 Applications(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06V40/23G06V20/53G06F18/2148
Inventor 朱超彭宇新
Owner PEKING UNIV
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