Affine-invariant Sketch-based Retrieval of Images 

INVESTIGATORS

Prof. Horace H S Ip, Mr. Angus Cheng
 
 
BRIEF DESCRIPTION
Digital library and multimedia database is one of the major application areas that based on various research techniques from image processing and computer vision in recent years. Using textual keyword for browsing or searching of visual data has been shown inefficient. Therefore, the feasibility of exploiting image content directly as a key has been widely studied in recent years. However, this approach is sensitive to certain transformations. These transformations included translation, rotation, scaling and affinity. In this research, we aim to develop techniques that based on hand-drawn sketch to retrieve the images, say top ten resemblance, that similar to the input sketch. So as to combat difficulties of certain transformations, shape descriptor from both images and sketches will be complemented with a variety of invariant factors.
 
 
FUNDING AGENCY
This work is part-supported by a HKSAR RGC grant No: RGC CityU 1113/99E (9040437).
 
 
PUBLICATIONS
  1. Horace HS Ip, Angus KY Cheng, William YF Wong, "Affine-invariant Sketch-based Retrieval of Images", CGI Conference 2001, July, 2001.

Sketch by User

Image Query Result