제어로봇시스템학회:학술대회논문집
- 1986.10a
- /
- Pages.331-336
- /
- 1986
The rocognition of two-dimensional objects using the inverse histogram
인버스 히스토그램을 이용한 다수의 이차원 물체 인식
Abstract
Because the threshold technique using the histogram of intensity is the most attractive for segmentation in the sense of fast image processing, this paper defined the new function of inverse histogram of intensity and found out a threshold by means of it. The segmented errors are removed by regulating a scan size of blob coloring. Blob-coloring algorithm presented by [6] was reproved for good performance i.e., no change of feature in bolobs after blob coloring. The ratio of successful recognition was about 85 percents.
Keywords