Applying a novel overlap splitting method and wavelet transformation to vehicle detection

Yi Nung Chung, Ming Tsung Yeh, Tun Chang Lu, Chao Hsing Hsu, Yu Xian Huang

Research output: Contribution to journalArticlepeer-review


The vehicle detection and tracking on urban environment is more complicated. In general, the system needs additional image processing to eliminate excess noise. The shadow and vehicle overlapping situation are usually the major interference for object counting and classification. In this paper, it proposes to use the color space adjustment for transforming the RGB image to HSV color space which is easy to reduce the shadow effect after using the threshold background subtraction to obtain the foreground objects. Afterward the system applies the edge-confined wavelet enhancement filter to improve the edge and boundary of detected vehicles, which assists to extract the features of object more accurately. Moreover, a novel method denoted as soft medium separation algorithm, which is used to separate the overlapped vehicles if the border width of detected object is greater than the compared threshold is designed in this paper. As the experimental analysis, that proves the vehicle detection using the proposed method is more accurate.

Original languageEnglish
Pages (from-to)2873-2878
Number of pages6
JournalICIC Express Letters
Issue number10
Publication statusPublished - 2015 Jan 1

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Applying a novel overlap splitting method and wavelet transformation to vehicle detection'. Together they form a unique fingerprint.

Cite this