Image Indexing
Image Indexing was a project that I created as an Independent Study course during the Fall 2007 semester. I was initially fascinated by image processing and its applications searching and video processing, and later decided to pursue an Independent Study under Professor Jie Wei. This project was rather enjoyable as I was able to use my own ingenuity to experiment with different algorithms for texture mapping.
The project is a search engine that allows a user to query a database of images based on different cues. The cues consist of textual information, such as a title or keyword set; and visual information, such as an image's histogram, dominant color, and texture. A querying user was able to suggest changes to the associated textual data, where the administrator would need to accept or deny the proposed modifications.
The software incorporated my own custom parser for a .HIST file, which was able to read and write the histogram information for a grayscale input image. I also used the .Net framework objects to expedite the development process by providing many ways of handling image data. The application was in a Multiple Document Interface (MDI) format, with user authentication, and an administrative interface for populating the database and
Image Indexing was built using C#, subversion, and MySQL.