30 October 2009 A region labeling algorithm based on block
Author Affiliations +
Proceedings Volume 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis; 74950M (2009) https://doi.org/10.1117/12.831249
Event: Sixth International Symposium on Multispectral Image Processing and Pattern Recognition, 2009, Yichang, China
The time performance of region labeling algorithm is important for image process. However, common region labeling algorithms cannot meet the requirements of real-time image processing. In this paper, a technique using block to record the connective area is proposed. By this technique, connective closure and information related to the target can be computed during a one-time image scan. It records the edge pixel's coordinate, including outer side edges and inner side edges, as well as the label, and then it can calculate connecting area's shape center, area and gray. Compared to others, this block based region labeling algorithm is more efficient. It can well meet the time requirements of real-time processing. Experiment results also validate the correctness and efficiency of the algorithm. Experiment results show that it can detect any connecting areas in binary images, which contains various complex and quaint patterns. The block labeling algorithm is used in a real-time image processing program now.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Jing Wang, Jing Wang, } "A region labeling algorithm based on block", Proc. SPIE 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis, 74950M (30 October 2009); doi: 10.1117/12.831249; https://doi.org/10.1117/12.831249


Back to Top