Color-content based digital photo sorting and presentation

• Main Author: Li, Yejing
• Other Authors: Tan Yap Peng
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Computer science and engineering::Computer applications::Arts and humanities; DRNTU::Visual arts and music::Photography; DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/61498
• Institution: Nanyang Technological University

This project is a smart solution for color-content based digital image sorting and presentation application. It manages digital images across different software platforms based on a sequence of sustainable algorithms. Moreover, it emphasizes the importance and impact of color as a sorting tool from a psychological point of view. At the same time, it is inspired and motivated by the popular use of online photo sharing and management platforms, as well as the inadequacy of color-content reference in traditional sorting methods. With this application, images in a designated location can be analyzed and sorted based on the color information. More specifically, the dominant color composition data are generated from MATLAB image processing with various data partitioning technics including minimum variation quantization and k-means clustering based on RGB and CIELAB color spaces. In order to sort and present the images, a sequence of algorithms are to discuss the dominance of color, color distance, color similarity, image similarity of dominant color composition, image connectivity and optimized presentation route planning. Further more, the analysis outcomes can be extracted and visualized in the browser space with HTML and JavaScript. To sum up, this is a collaboration project of different areas of study and platforms. It aims to provide an efficient solution for color-content based digital image management, yet it is also a framework for discussion and further implementation with integrated algorithms that provide various feature analyses.