13 - Grading of Potatoes

Abstract

Grading and sorting of potatoes ensure that derived products meet the defined grade requirements for sellers and the expected quality for buyers. Grading is particularly important for potatoes because the size, shape, color, and defects depend greatly on environmental conditions and handling and is performed primarily by trained human inspectors who assess the potatoes by seeing or feeling a particular quality attribute. However, there are some disadvantages to using human inspectors, including inconsistency, short supply of labor, and the expense of the large amounts of time required due to the huge volume of production. Product experts characterize potato defects and diseases based on color and shape features, and thus computer vision may improve inspection results and be able to take over the visually intensive inspection work from human inspectors. Automation is desirable because it can ensure consistency in product quality and can handle large volumes. A completely automated inspection station requires the incorporation of machine vision and automation into a system consisting of the appropriate hardware and software for product handling and grading. Factors such as size, shape, greening, cracks, scabs, etc. determine the final grade of a potato. Some researchers have implemented algorithms to detect potato features such as bruises and have shown that Kohonen's self-organizing map is suitable for identifying both bruised and green areas on potato flesh. As bruises clearly contrast with healthy potato flesh, which is very uniform in color, excellent results should be easily obtained. Further developments will involve improvement in image capture, measurement, and processing, as well as assessment of the relevant surfaces of healthy and bruised areas.