Implementation of Hough transform for fruit image segmentation

Abstract A computer vision system tries to mimic our primary sense (sight) in order to gather information without the need for physical interaction, in fact such systems are able to grade automatically, and extract useful information with a degree of sensitivity closer to that of a human, reducing considerably the margin of error. By performing digital image processing, defined as the acquisition and processing of visual information by computer, computer vision systems allow analyzing image data for specific applications in order to determine how images can be used to extract the required information. Among the most important features for accurate classification and sorting of products it can be mentioned the shape. The shape of objects or regions of interest are important features used for content representation, and require good segmentation to detect objects or regions. Basically, shape characterization is of two types: boundary-based and regionbased. Boundary-based shape features include rectilinear shapes, polygonal approximation, finite element models, and Fourier-based shape descriptors. Region-based features include statistical moments and grid-based approaches. Object shape detection using a technique based on Hough Transform for further segmentation is presented on this paper.