Connected component labeling of binary images on a mesh connected massively parallel processor
Computer Vision, Graphics, and Image Processing
Parallel Architectures and Algorithms for Image Component Labeling
IEEE Transactions on Pattern Analysis and Machine Intelligence
A systolic approach for real time connected component labeling
Computer Vision and Image Understanding
Sequential Operations in Digital Picture Processing
Journal of the ACM (JACM)
Connectivity in Digital Pictures
Journal of the ACM (JACM)
Computing connected components on parallel computers
Communications of the ACM
Digital Picture Processing
Linear-time connected-component labeling based on sequential local operations
Computer Vision and Image Understanding
A linear-time component-labeling algorithm using contour tracing technique
Computer Vision and Image Understanding
Hybrid object labelling in digital images
Machine Vision and Applications
Optimizing two-pass connected-component labeling algorithms
Pattern Analysis & Applications
Fast connected-component labeling
Pattern Recognition
Fast connected-component labelling in three-dimensional binary images based on iterative recursion
Computer Vision and Image Understanding
A Run-Based Two-Scan Labeling Algorithm
IEEE Transactions on Image Processing
Hi-index | 0.10 |
Label-equivalence-based connected-component labeling algorithms complete labeling in two or more raster scans. In the first scan, each foreground pixel is assigned a provisional label, and label equivalences between provisional labels are recorded. For doing this task, all conventional algorithms use the same mask that consists of four processed neighbor pixels to process every foreground pixel. This paper presents a simple yet efficient first-scan method for label-equivalence-based labeling algorithms. In our method, foreground pixels following a background pixel and those following a foreground pixel are processed in a different way. By use of this idea, the pixel followed by the current foreground pixel can be removed from the mask. In other words, the mask used in our method consists of three processed neighbor pixels. Thus, for processing a foreground pixel, the average number of times for checking the processed neighbor pixels in the first scan is reduced from 2.25 to 1.75. Because the current foreground pixel following a background pixel or a foreground pixel can be known without any additional computing cost, our method is efficient for any image that contains at least one foreground pixel. Experimental results demonstrated that our method is effective for improving the efficiency of label-equivalence-based labeling algorithms.