We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Detecting region of interest for cadastral images in Taiwan.
- Authors
Lin, Guo-Shiang; Tuan, Nguyen-Minh; Chen, Wen-Jan
- Abstract
In the paper, we proposed a coarse-to-fine scheme to automatically detect the regions of interest (ROIs) for digitized cadastral images in Taiwan. To consider some issues such as the skew effect, image quality, and noise in digitized cadastral images, the proposed scheme is composed of four parts: pre-processing, skew correction, noise reduction, and ROI localization. In the pre-processing, each cadastral image is normalized and the prior knowledge is used to find the candidate region of the ROI. To reduce the impact of noise and poor contrast on line detection, an adaptive thresholding is adopted. To decrease the skew effect, the detected horizontal and vertical lines are analyzed to estimate the skew angle. After skew correction, an adaptive noise reduction algorithm is devised to reduce the effect of marginal, artificial, and random noise. The coordinates of the candidate region in the de-skew image with high resolution can be found and then the ROI boundary can be located correctly in the fine detection. Experimental results demonstrate that the proposed scheme can effectively and correctly localize ROIs in digitized cadastral images.
- Subjects
TAIWAN; IMAGE quality in imaging systems; NOISE control; LOCALIZATION theory; ALGORITHMS; SOCIAL history
- Publication
Multimedia Tools & Applications, 2017, Vol 76, Issue 23, p25369
- ISSN
1380-7501
- Publication type
Article
- DOI
10.1007/s11042-017-4504-5