Our works on this topic focus on color edge detection, seeded region growing and seeded region
aggregation for salient object generation. Salient objects are needed to be the semantic objects from human point of view, but they should have the capability to indicate the presence or absence of the associated semantic image/video concepts.
2. Color Image Segmentation:
The color edges are first obtained by using our improved color edge detection algorithm, and then these connected edges are merged. The centroids of two neighboring color edges are then used as seeds for region growing. This is the first work to automate the seeded region growing technique. By involving the semantic knowledge about the image components in the seed selection procedure, this technique is also very attractive for salient object extraction via a seeded region aggregation procedure.
3. Confidence-Assisted Salient Object Detection:
4. Integrate Camera Metadata for Attended Object Detection: