Smart Detection Of Mud In Lotus Root Using Machine Vision Technologies

Lotus root is a kind of aquatic plant cherished by people in the Asian continent and some other parts of the world.During its growth,harvesting,and cleaning,the lotus root holes may enter the silt which will make it unsafe for consumption.Before using lotus root to prepare food,it is necessary to inspect the lotus root holes to ensure its hygiene and safety.The traditional detection method is used to manually inspect and detect the presence mud in the lotus root hole.This method is inefficient and easy to miss.In order to adapt to the large-scale production of lotus root,it was necessary to improve the manual quality inspection process and realize the automatic detection of lotus root hole.The use of machine vision system and technology for inspection and detection is both high-speed and efficient,and also has good consistency.This paper proposed lotus root mud hole inspection algorithms,i.e.,edge detection,adaptive threshold and global threshold segmentation,and region growth segmentation to detect whether there is mud in the holes of the lotus root.The detection algorithm uses the differential pixel intensity gradients of the lotus root and the mud as the parameter to analyze the quality of the lotus root images.A lotus root hole inspection system was setup to carry out the detection of the mud.Before the lotus root went through the inspection process,browning prevention control was carried out on the lotus root.The polyphenol in the lotus root causes the lotus to turn brown by oxidation process if not treated.This was done to prevent the algorithm from mistakenly identifying the browned parts as mud.The hardware components used in this study were the light source,the camera,the illumination box,and the computer.Then the software components are the lotus root mud hole detection algorithm proposed in this paper.A ring white coaxial LED light was used as illumination source to enhance the lotus root image before acquisition by the camera.The illumination box was darkened to prevent light reflections and the images from being shadowy.After the image acquisition,the developed detection algorithms were then employed to inspect and detect the lotus root hole for the presence of mud mainly by their differential pixel intensities.After experiments had been carried out,the experiments results showed that the lotus root hole inspection algorithms proposed in this paper had effectively detected mud in lotus root holes at a faster speed,more efficient and accurate rate,and all at a lower cost than traditional manual quality inspection method.