A comparative approach for image segmentation to identify the defected portion of apple

A crucially significant process for the automatic fruit grading system is image segmentation. The area of interest is extracted by separating the image into several areas. Intensity of the object color surface varies with the illumination. The external fruit properties namely shape, size, texture and color, give base to various fruit quality detection technique. As it is very difficult to sort out the good quality fruits, so the aim of machine based system is to replace manual techniques. With a large demand of fruits the ineffective manual monitoring poses a problem. Various methods such as Otsu, k-means, fuzzy c-means and watershed segmentation are used for image segmentation. Speeded up robust technique estimate the local features. Implementation of all segmentation methods on fruit images is applied and a comparative research outcome is projected to find the defected portion of fruits.