Pedestrian target verification algorithm and system based on recursive structure of vehicle-mounted monocular camera

A monocular camera and pedestrian target technology, which is applied in the field of vehicle-assisted driving, can solve the problems of false detection of pedestrian targets, achieve the effects of reducing the number of false detection targets, good pedestrian target detection results, and improving timeliness

Active Publication Date: 2021-11-05
DALIAN NATIONALITIES UNIVERSITY
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem of false detection of pedestrian targets in the vehicle-mounted monocular camera pedestrian detection, the present invention provides a pedestrian target verification algorithm and system based on a recursive structure of the vehicle-mounted monocular camera

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
  • Pedestrian target verification algorithm and system based on recursive structure of vehicle-mounted monocular camera
  • Pedestrian target verification algorithm and system based on recursive structure of vehicle-mounted monocular camera
  • Pedestrian target verification algorithm and system based on recursive structure of vehicle-mounted monocular camera

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0076] Error detection of pedestrian calibration frame in outdoor single-person environment

[0077] This embodiment aims at the phenomenon that the output pedestrian target frame is incorrectly detected due to the complexity of the feature information of the image to be detected in the outdoor single-person environment, and the present invention is applied to figure 2 In the case of the image to be detected as input to the pedestrian detection system, image 3 is the output result of the wrongly detected pedestrian target box when the recursive operation is not started, Figure 4 It is the output result of successfully removing the false detection pedestrian target frame after the recursive calculation verification.

Embodiment 2

[0079] Pedestrian calibration frame error detection in indoor single person situation

[0080] This embodiment aims at the phenomenon of misdetection of the output pedestrian target frame due to the complex feature information of the image to be detected in the case of a single person indoors, and applies the present invention to Figure 5 In the case of the image to be detected as input to the pedestrian detection system, Image 6 is the output result of the wrongly detected pedestrian target box when the recursive operation is not started, Figure 7 It is the output result of successfully removing the false detection pedestrian target frame after the recursive calculation verification.

Embodiment 3

[0082] Misdetection of pedestrian calibration frame in outdoor multi-person environment

[0083] This embodiment aims at the phenomenon that the output pedestrian target frame is misdetected due to the complexity of the feature information of the actual image to be detected in the outdoor multi-person environment, and applies the present invention to Figure 8 In the case of the image to be detected as input to the pedestrian detection system, Figure 9 is the output result of the wrongly detected pedestrian target box when the recursive operation is not started, Figure 10 It is the output result of successfully removing the false detection pedestrian target frame after the recursive calculation verification.

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

Based on the recursive structure vehicle-mounted monocular camera pedestrian target verification algorithm and system, it belongs to the field of vehicle-assisted driving technology and is used to solve the problem of pedestrian target false detection in the vehicle-mounted monocular camera pedestrian detection. The main point is to process the pedestrian target image As a new image to be detected, it is re-input into the HOG feature pedestrian detection system F, and the output pedestrian detection frame coordinates and the corresponding pedestrian target image are obtained. The effect is to apply the recursive structure to the output result of the HOG feature pedestrian detection system. The output result of the sharpening process is re-input into the pedestrian detection system as feedback, without changing the structure of the existing HOG feature pedestrian detector, and it can be used in pedestrian detection due to low algorithm accuracy, too complex input image scene and irrelevant characteristic information The wrongly detected pedestrian calibration frame caused by factors such as the pedestrian detection system is effectively removed, and the accuracy of the pedestrian detection system is improved.

Description

technical field [0001] The invention belongs to the technical field of vehicle-mounted assisted driving, and in particular relates to a pedestrian target verification algorithm based on a recursive structure. Background technique [0002] With the advent of the era of unmanned vehicles, the processing technology related to pedestrian detection has become one of the hot research directions. Pedestrian detectors based on HOG features have high computational accuracy and low complexity and become a popular tool in machine learning. There are a large number of this type of detectors, and the structure is easy to understand. However, due to a series of factors such as different scene differences, uneven light, and dense occlusions, the complexity of the detection information of the image to be detected will increase, and it is easy to make the pedestrian's characteristic information False detection occurs, reducing the accuracy of the detection system. The pedestrian detection ...

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): G06K9/00
CPCG06V40/10
Inventor 毛琳杨大伟陈思宇张静
Owner DALIAN NATIONALITIES 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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap