الفهرس 
contentsOnly 14 pages are availabe for public view
 |  | from | 149 |  |  |
| | | from | 149 | | |
Abstract
Segmentation plays an important role in digital media processing, pattern recognition, and computer vision. The task of image/video segmentation emerges in many application areas, such as image interpretations, video analysis and understanding, video summarization and indexing, and digital entertainment. Over the last two decades, the problem of segmenting image/video data has become an important area and had significant impact on both new pattern recognition algorithms and applications.
There are other possible applications, such as medical diagnosis, tele-education, industrial inspection, environmental monitoring.
The objective of this thesis is to develop image and video segmentation framework based on the new filed of fast discrete curvelet transform (FDCT) for different images and video sequences. The thesis is divided into two parts of segmentation techniques. One is image segmentation based on curvelet transform; another is video segmentation based on curvelet transform.
In the first part of this thesis, image segmentation based on the second generation of the curvelet transform is proposed. Canny edge detector is used in conjunction with the curvelet transform to initially obtain the object edges. in the image then the result is processed using some morphological filters to obtain the accurate object. Simulation results show that the curvelet transform can help to detect accurate edges of the object to be accurately extracted.
In the second part of this thesis, video segmentation based on second generation of the curvelet transform is proposed. The initial video frames are processed using Canny edge detector in the curvelet domain and collected using OR logical operator to represent mask1. Also, initial frames of the video sequence are used to obtain another mask by subtracting initial frames from the first frame and the resulted frames are averaged to obtain mask2. The AND operator is used to collect mask1 and mask2 to obtain final mask3. The obtained mask3 is used to estimate the background. Each frame in the video sequence is subtracted from the estimated background to obtain the moving object. Simulation results show fantastic extraction of the moving object compared with other methods.