Forward/backward contour tracing with feedback
IEEE Transactions on Pattern Analysis and Machine Intelligence
A contour detection algorithm based on the minimum radial inertia (MRI) criterion
Computer Vision, Graphics, and Image Processing
A Three-Module Strategy for Edge Detection
IEEE Transactions on Pattern Analysis and Machine Intelligence
IEEE Transactions on Pattern Analysis and Machine Intelligence
Edge contours using multiple scales
Computer Vision, Graphics, and Image Processing
Variable-resolution character thinning
Pattern Recognition Letters
Edge and Curve Detection: Further Experiments
IEEE Transactions on Computers
Extracting and Labeling Boundary Segments in Natural Scenes
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.10 |
Computer vision algorithms have relied primarily on either uniform resolution or multi-resolution techniques. In this paper, the question we try to answer is: How can varying resolution spatially within an image help in extracting information out of pictures? Specifically. we look at the problem of boundary following. To deal with these problems we design variable resolution (VR) masks, whose centers are windows in normal resolution, with each of the peripheral cells used to keep some information at a reduced resolution. In this way, VR approaches effectively look at a large region of the original image at a cost only slightly higher than processing a small region using uniform resolution schemes. The proposed algorithms are demonstrated to perform better than other well-known methods.