Perceptual image processing algorithm developing

• Main Author: Htoo, Maung Maung
• Other Authors: Lin Weisi
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Computer science and engineering::Computer applications
• Online Access: http://hdl.handle.net/10356/59074
• Institution: Nanyang Technological University

Digital images have one of the most familiar things we interact in our daily life because we can take photos anywhere anytime with our personal mobile phone. Therefore, handling the digital images becomes popular. In the past decades, researchers have been developed various algorithms for image-related tasks but most of the algorithms still need the manual labor to accomplish the tasks. This project aims to develop useful and efficient application to automate some of the tasks such as select, edit, organize and other image handling tasks. This primary object is to automate the scoring process of digital images without human help. To save manual works, blind image quality assessment method has been chosen as the method to get the perceptual score of the image. This method is then integrated together with the graphical user interface for the convenience use. Through this project, risk-driven process model, spiral project model has been chosen as project development model since it is able to cope with the changes that software development generally entails. The entire project is separated into several development phases. Goals are set for each phase and perform the analysis at the end of each phase. The changes and modification have been done in the next phase based on the previous analysis. After development of the different versions of the application, the results are applicable and useful for personal use. The application can not only calculate the perceptual image score, but also offer other functionalities such as ranking and online image search. This project will be continue to work and offer more capability for the automation of the digital image handling. The possible future works to continue are making the application to be able to run cross platform (not only on the window, but also on Macs and Linux OS). Since the smart phone usages are rapidly growing mobile application could also have been developed based on this project