• Journal of Korea Multimedia Society
• /
• v.14 no.5
• /
• pp.622-631
• /
• 2011

Corners play an important role in describing object features for pattern recognition and identification. This paper proposed a fast and adaptive corner detector in both coarse and fine scale, followed by the framework of the curvature scale space (CSS). An adaptive curvature threshold and evaluating of angles of corner candidates are added to original CSS to remove round corners and false corners in the detecting process. The efficiency of proposed method is compared to other popular detectors in both accuracy criteria, stability and time consuming. Results illustrate that the proposed method performs extremely surpass in both areas.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.27 no.1
• /
• pp.37-55
• /
• 2023

In this paper we propose a new algorithm for detecting and counting flowers in a complex background based on digital images. The algorithm mainly includes the following parts: edge contour extraction of flowers, edge contour determination of overlapped flowers and flower counting. We use a contour detection technique in Computer Vision (CV) to extract the edge contours of flowers and propose an improved algorithm with a concave point detection technique to find accurate segmentation for overlapped flowers. In this process, we first use the polygon approximation to smooth edge contours and then adopt the second-order central moments to fit ellipse contours to determine whether edge contours overlap. To obtain accurate segmentation points, we calculate the curvature of each pixel point on the edge contours with an improved Curvature Scale Space (CSS) corner detector. Finally, we successively give three adaptive judgment criteria to detect and count flowers accurately and automatically. Both experimental results and the proposed evaluation indicators reveal that the proposed algorithm is more efficient for flower counting.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.409-411
• /
• 2001

임의의 입력 영상을 이해함에 있어서 코너점은 디지털 영상의 중요한 정보가 집중되어 있기 때문에 형태를 분석하는데 있어 중요한 요소이다. 본 논문은 영상의 중요한 정보 요소인 코너점을 보다 정확하게 추출하기 위하여 Farzin Mokhtarian과 Riku Suomela가 제안한 CSS(Curvature Scale Space) 방법에 기초한 다면체 모델 방법을 이용한 새로운 알고리즘을 제안하고자 한다.