<![CDATA[Development in the field of image processing had been a lot of inventing new technologies. Applications have been developed are character recognition, fingerprint recognition, face recognition, industrial applications, and answer sheet processing. Answer sheet processing applications using a scanner has been widely applied. However, processing the answer sheets using scanners requires special paper. It has been developed processing the answer sheets using a CCD camera. The CCD optical sensors are categorized as the price is quite expensive compared with CMOS. This research aims to obtain the performance of thresholding and region merging in multiple-choice answer with 1.3 mega pixel web camera.The web camera utilize CMOS optical sensor. Thresholding algorithm is used to convert from black-and-white image into a binary image. Thresholding operation classifies the degree of gray value of each pixel into two classes, black and white. Region merging algorithm is a process to incorporate the region that the distance is less than the threshold. Neighboring areas with the most minimum distance are combined. This process continues until the minimum distance between the neighboring region is greater than the threshold.1.3 mega pixels web camera has a low uniformity and also suffers in low light condition.This limitation can be overcome by adding a tolerance on the thresholding process. The system uses Delphi 7programming languages on Windows XP operating system. This research shows that a web camera with alow level of uniformity and suffers in low light condition can retrieve the image data from answer-sheet formwith a precision rate of 96.96%.]]>
Copyrights © 2012
