Interactive mobile information hub

• Main Author: Yang, Zhiyong
• Other Authors: Yap Kim Hui
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/60593
• Institution: Nanyang Technological University

The number of smartphones have been increasing over the years. There are currently 78% of people owning at least one smartphone in Singapore as of June2013 up from 74% in the previous year[1]. Many applications are being developed on various platforms and android is the most popular platform in recent years having captured 79% market share last year[2]. Other than the ever increasing smartphone users, there has been a trend to digitalize hard copy documents into softcopy files which can be achieved by Optical Character Recognition(OCR) technologies. Therefore, the author has integrated OCR function available in the compiled android native library into Mobile Visual Search(MVS) application such that it will provide as an additional function on top of the current MVS application to do a text detection on the image and return result to the user before sending the query image to the server for recognition and retrieval of results which can be done in an offline manner. The process of integrating OCR into MVS application will be explained in Chapter 5.4. In this project, the author has researched on the various OCR technologies to select the best performing OCR engine to be used in this project. The author has also study the pre-processing and post-processing and has provided a solution to eliminate unwanted result from the OCR output to achieve a better OCR result Also, a detailed analysis was done on the challenges of OCR where the author has studied the effects of the environment affecting the quality of the input images. The author found that the input image has a direct impact on the output of the OCR such as the lighting condition, scale, rotation of the image and font of the building.