Research On Mango Intelligent Detection And Classification System Based On Pipeline

With the continuous progress of industrialization in China,agricultural products have gradually moved from high yield to high quality.In recent years,mango sales in the overseas market are increasing,mango products in the overseas market prospects are very good.After picking in China,mango is generally classified manually,i.e.simply according to the size and appearance of the fruit.Due to the inaccurate classification of mango and the excessive input of manual classification,the product value is lost and the profit is meager.Therefore,it is of great significance to improve the classification of Postharvest mango in China.This paper studies the intelligent detection and classification system of mango on the assembly line.Mango shape detection algorithm based on pixel depth and computer image processing technology is used to accurately calculate the color parameter,shape parameter,volume value and defect area of mango.The results showed that the relative errors of mango volume and defect area were less than 2.0% and 2.0%,respectively.According to the characteristics of mango such as color,shape,volume and defect,this paper uses BP neural network algorithm and support vector machine algorithm to classify mango.The experimental results show that the algorithm based on the improved granular support vector machine(GSVM)is efficient,accurate and simple.In the process of intelligent detection and classification of mango production line,the quality of product classification is improved and the development of agricultural product processing industry is promoted.