Group 7 | Tabiscope : Mobile Device Camera Connector

Project Goal and Relevance The goal of this project is to provide a low-cost alternative to the traditional endoscopic image systems that are expensive and not portable. There are already existing solutions such as Endoscope-I that consist of an imaging software and an endoscope adapter to iPhone. Clinicians who used it, however, pointed out some limitations of this existing solution. Firstly, they complained that Endoscope-I’s resolution is limited to small screen size of iPhone. To address this issue, our group chose to use an Android tablet for bigger screen size. Secondly, its image quality is not as good as traditional imaging towers’. We suspect that it is because Endoscope-I’s iPhone application uses iPhone’s existing internal camera application and there is no image processing, whereas traditional imaging towers usually have image processing processors. Therefore, a camera application with low control and real-time image process algorithm need to be implemented on the tablet to address these issues. Since a tablet has stronger computational power than an iPhone, some real-time image processing can be done to improve the image quality. For example, using Randomized Hough Transformation (RHT), an algorithm can be implemented for a tablet automatically to detect the region of interest, usually a circular shape, and for the tablet to focus on this area. More specifically, among many issues with image quality, clinicians pointed out that the specular reflection phenomena occurs more often in iPhones than it does in the towers. This issue might get resolved by implementing image processing algorithm for removing specular reflection. Although there is no known perfect algorithm that can remove specular reflection from image, we review these two papers, Stehle (2006) and Kaup et al. (2005), in order to assess whether it is feasible to provide a software-based solution to the existing issues with specular reflection.