A New Iterated Connected Components Labeling Algorithm Based on Medical Segmentation

Connected Component labeling of a binary image is an important task especially when it is used in medical images for recognition purposes. This research is an advance step for applying a proposed algorithm for allocating connected components labeling of medical images. We explore and simulate iterative method towards the development of an automated system for the purpose of connected components labeling to be applied for constructing such labeling on colored images by repetition labeling on subimages which are originally segmented from the original image. These sub-images are formed from the whole image segmentation process. The algorithm is a simulation process on colored images for practical medical image. Two process algorithm is applied for labeling. The first process is row-wise from left to right. The second one is column-wise, from top to bottom. One important application is the medical image simulation that shows an interesting topic related to heart failure and heart attack. The efficiency of the proposed algorithm is noticed compared with the conventional algorithms in term of memory space and accuracy as well. For the small size images 128 X 128, the iteration version was acceptable in term of computer time and it is recommended to be used to label specific medical images. We experimented 2 sizes: 128 X 128, 256 X 256 and The speed-up was obtained using 128 X 128 size, using all 8connected neighbors instead of looking at 4 immediate neighbors, which yields to reduction in the number of iterations required with a little increase in search time. Index Term— Segmentation, sub-connected component labeling, medical images, threshold image processing.