Detection of Lotus Root Contaminants Using Intelligent Visual Machine Vision Techniques

Abstract

Lotus root, which is the most important aquatic vegetable in China, is manually inspected for quality by experts to detect impurities in a food production plant in China. There is therefore the need to update this inspection process in order to improve the quality and safety of lotus root. Machine vision systems and techniques are used for consistent, efficient, effective, and reliable inspection of images. This technology has helped several industries in their effort to visually inspect and analyze products. The lotus root inspection system has been proposed to inspect the lotus roots for impurities. The detection algorithms use the size, shape, texture and color of the lotus root and impurities images as parameters to analyze the quality of lotus roots. The lotus root undergoes some cleaning processes before image acquisition and image processing. The camera and illumination used, in collaboration with the edge detection, and image segmentation techniques, efficiently and effectively exposed the impurities in the lotus root at a much faster rate. This proposed inspection technique is less expensive, effective and more efficient compared to the traditional human inspections.