Texture feature performance for image segmentation
Pattern Recognition
On-Line and Off-Line Handwriting Recognition: A Comprehensive Survey
IEEE Transactions on Pattern Analysis and Machine Intelligence
Class-Specific, Top-Down Segmentation
ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part II
Combining Top-Down and Bottom-Up Segmentation
CVPRW '04 Proceedings of the 2004 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'04) Volume 4 - Volume 04
Detecting Text Lines in Handwritten Documents
ICPR '06 Proceedings of the 18th International Conference on Pattern Recognition - Volume 02
| Hi-index | 0.00 |
This paper compares and analyzes different Segmentation techniques that are used in pattern analysis and machine intelligence. Comparing motion segmentation, script-independent text line segmentation in freestyle handwritten documents and combined top-down and bottom up segmentation based on density estimation and state of the art image segmentation technique. It presents experimental results and quantitative evaluation, demonstrating the resulting approach is effective for very challenging data. The main novel aspects of this work are the fragment learning phase, which efficiently learns the figure-ground labeling of segmentation fragments, even in training sets with high object and background variability; combining the resulting top-down segmentation with bottom-up criteria; and the use of segmentation to improve recognition.