Simple colour image segmentation of bicolour food products for quality measurement

This paper introduces a colour image segmentation algorithm for bicolour food products. The algorithm has been developed for colour quality measurement in manufacturing processes. a∗ and b∗ components of CIE L∗a∗b∗ colour space are sampled from pixels in the food product region in an image. A polynomial equation is obtained from this colour data by using a least square fitting method. All pixels in the image are tested with this equation, and pixels having colour values close to the polynomial equation are regarded as ones in the food product region in the image. This algorithm is useful when multiple products are captured in an image and the food region needs to be segmented. The procedures of the algorithm and its application using French fried chips and dried apple slices are presented.