Determining the accuracy of a proof sheet using image segmentation and edge detection techniques

Abstract

This research proposes a technique for next proofing in packaging printing process based on Boundary Description, Absolute Chain Code and Pixel Line Length Determination. The research started with creation of an input file operation which was the first stage of an algorithm. A set of 26 English alphabet characters with the font type of Cordia New having a 16 point font size was then scanned and used as a prototype. Each character in the alphabet was made a function using the code extracted. The algorithm was used to translate the characters in the scanned document. A Matching Algorithm would then detect the accuracy between the printed sheet and the original file. The algorithm was limited to be used with an image file of .bmp format and a bilevel color image specification. This algorithm was fast and very easy to operate. It was flexible to apply to other kinds of alphabet characters. The efficiency of the algorithm depended on the quality of the apparatus such as printer and scanner.