DocumentCode :
3263821
Title :
Brick path detection from shape pattern and texture feature
Author :
Woo, Byung-Seok ; Yang, Sung-Min ; Jo, Kang-Hyun
Author_Institution :
Sch. of Electr. Eng., Univ. of Ulsan, Ulsan, South Korea
fYear :
2011
fDate :
20-22 Dec. 2011
Firstpage :
78
Lastpage :
83
Abstract :
Pedestrians are involved in half of the traffic accidents according to the statistics. It shows that the safety of pedestrian passage is so important. In this paper, a detection method for paved brick path is proposed which is based on the analysis of shape pattern and texture feature in outdoor environment. In image, a brick appears to a regular type because of its specific shape which repeats with its sized gap. However, the type of brick path varies according to the surrounding environment or section of use for the handicapped people. This paper shows the method which extracts the shape pattern by analyzing single surface polygon and its frequency appearing in the road area. The shape pattern is used to detect similar shape regions. Some regions aren´t detected because the extraneous substances or chopped bricks distort the original shape. This kind of problem can be solved by texture feature analysis. Analyzed values of texture feature in previously detected regions yield Gaussian distribution. An analyzed value of texture feature in each undetected region also is extracted and then the value is checked it is satisfied with Gaussian distribution or not. Finally the satisfied regions are detected as brick path. The experimental result is analyzed that true positive is 97%.
Keywords :
Gaussian distribution; computational geometry; feature extraction; handicapped aids; image texture; object detection; shape recognition; Gaussian distribution; brick path detection; chopped brick; extraneous substance; paved brick path; shape pattern; single surface polygon; texture feature; Azimuth; Feature extraction; Gaussian distribution; Image edge detection; Roads; Shape; Silicon;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
System Integration (SII), 2011 IEEE/SICE International Symposium on
Conference_Location :
Kyoto
Print_ISBN :
978-1-4577-1523-5
Type :
conf
DOI :
10.1109/SII.2011.6147423
Filename :
6147423
Link To Document :
بازگشت