Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature

Weilong Song; Guangming Xiong; Linlang Cao; Yanhua Jiang

doi:10.1007/978-3-642-22418-8_69

Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature

Weilong Song^*, Guangming Xiong, Linlang Cao, Yanhua Jiang

^*Corresponding author for this work

School of Mechanical Engineering

Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Citations (Scopus)

Abstract

Object detection using image processing mainly includes single-camera based detection and stereo vision based detection. Stereo vision-based object detection approaches need to find correspondences from the left to the right image. In this paper, we proposed to detect the region of interest (ROI) of objects like cars, people, bikes using HOG descriptors and make correspondence matching only in ROI. Moreover, we used 3D reprojection method to calculate the depth and employed the subpixel disparity technology to increase the depth range resolution. By the proposed approach, the corresponding is only needed to be done for special features in the special region which improves the accuracy.

Original language	English
Title of host publication	Advances in Information Technology and Education - International Conference, CSE 2011, Proceedings
Pages	489-494
Number of pages	6
Edition	PART 1
DOIs	https://doi.org/10.1007/978-3-642-22418-8_69
Publication status	Published - 2011
Event	2011 International Conference on Computer Science and Education, CSE 2011 - Qingdao, China Duration: 9 Jul 2011 → 10 Jul 2011

Publication series

Name	Communications in Computer and Information Science
Number	PART 1
Volume	201 CCIS
ISSN (Print)	1865-0929

Conference

Conference	2011 International Conference on Computer Science and Education, CSE 2011
Country/Territory	China
City	Qingdao
Period	9/07/11 → 10/07/11

Keywords

Stereo vision
depth calculation
subpixel disparity

Access to Document

10.1007/978-3-642-22418-8_69

Cite this

Song, W., Xiong, G., Cao, L., & Jiang, Y. (2011). Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature. In Advances in Information Technology and Education - International Conference, CSE 2011, Proceedings (PART 1 ed., pp. 489-494). (Communications in Computer and Information Science; Vol. 201 CCIS, No. PART 1). https://doi.org/10.1007/978-3-642-22418-8_69

@inproceedings{4220d9c3209a4544814d6ec5b6c4a28b,

title = "Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature",

abstract = "Object detection using image processing mainly includes single-camera based detection and stereo vision based detection. Stereo vision-based object detection approaches need to find correspondences from the left to the right image. In this paper, we proposed to detect the region of interest (ROI) of objects like cars, people, bikes using HOG descriptors and make correspondence matching only in ROI. Moreover, we used 3D reprojection method to calculate the depth and employed the subpixel disparity technology to increase the depth range resolution. By the proposed approach, the corresponding is only needed to be done for special features in the special region which improves the accuracy.",

keywords = "Stereo vision, depth calculation, subpixel disparity",

author = "Weilong Song and Guangming Xiong and Linlang Cao and Yanhua Jiang",

year = "2011",

doi = "10.1007/978-3-642-22418-8_69",

language = "English",

isbn = "9783642224171",

series = "Communications in Computer and Information Science",

number = "PART 1",

pages = "489--494",

booktitle = "Advances in Information Technology and Education - International Conference, CSE 2011, Proceedings",

edition = "PART 1",

note = "2011 International Conference on Computer Science and Education, CSE 2011 ; Conference date: 09-07-2011 Through 10-07-2011",

}

Song, W, Xiong, G, Cao, L & Jiang, Y 2011, Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature. in Advances in Information Technology and Education - International Conference, CSE 2011, Proceedings. PART 1 edn, Communications in Computer and Information Science, no. PART 1, vol. 201 CCIS, pp. 489-494, 2011 International Conference on Computer Science and Education, CSE 2011, Qingdao, China, 9/07/11. https://doi.org/10.1007/978-3-642-22418-8_69

Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature. / Song, Weilong; Xiong, Guangming; Cao, Linlang et al.
Advances in Information Technology and Education - International Conference, CSE 2011, Proceedings. PART 1. ed. 2011. p. 489-494 (Communications in Computer and Information Science; Vol. 201 CCIS, No. PART 1).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature

AU - Song, Weilong

AU - Xiong, Guangming

AU - Cao, Linlang

AU - Jiang, Yanhua

PY - 2011

Y1 - 2011

N2 - Object detection using image processing mainly includes single-camera based detection and stereo vision based detection. Stereo vision-based object detection approaches need to find correspondences from the left to the right image. In this paper, we proposed to detect the region of interest (ROI) of objects like cars, people, bikes using HOG descriptors and make correspondence matching only in ROI. Moreover, we used 3D reprojection method to calculate the depth and employed the subpixel disparity technology to increase the depth range resolution. By the proposed approach, the corresponding is only needed to be done for special features in the special region which improves the accuracy.

AB - Object detection using image processing mainly includes single-camera based detection and stereo vision based detection. Stereo vision-based object detection approaches need to find correspondences from the left to the right image. In this paper, we proposed to detect the region of interest (ROI) of objects like cars, people, bikes using HOG descriptors and make correspondence matching only in ROI. Moreover, we used 3D reprojection method to calculate the depth and employed the subpixel disparity technology to increase the depth range resolution. By the proposed approach, the corresponding is only needed to be done for special features in the special region which improves the accuracy.

KW - Stereo vision

KW - depth calculation

KW - subpixel disparity

UR - http://www.scopus.com/inward/record.url?scp=79960400111&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-22418-8_69

DO - 10.1007/978-3-642-22418-8_69

M3 - Conference contribution

AN - SCOPUS:79960400111

SN - 9783642224171

T3 - Communications in Computer and Information Science

SP - 489

EP - 494

BT - Advances in Information Technology and Education - International Conference, CSE 2011, Proceedings

T2 - 2011 International Conference on Computer Science and Education, CSE 2011

Y2 - 9 July 2011 through 10 July 2011

ER -

Depth calculation and object detection using stereo vision with subpixel disparity and HOG feature

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this