Design Of Peanut Seeds Automatic Recognition System Based On Machine Vision

Peanuts are one of the important oil crops in our country. How to improveselect efficiency and quality of peanut seeds, which has become one of theimportant topics in related research fields. The machine vision technologyintroduced to the peanut seed feature extraction. This can solved weakness of thetraditional method that have long testing cycle, low efficiency, accuracy and poor.This has provided a theoretical basis and laid a solid technical foundation for thepreferred peanut seeds.In this paper, we have in-depth study machine vision detection method ofdomestic and overseas. By studying LabVIEW and image processing algorithms,we have designed a system that to detect peanut seeds feature extraction. Thuspeanut seeds feature extract can efficient, rapid, and convenient. Furthermorethrough the feature extract data can achieve precise sorting peanut seeds. Themain contents are as follows:Firstly, research and analysis of the current situation at home and abroad formachine vision applications in crops. Comprehensive advantages of variousmethods to determine the peanut seeds feature extraction in overall design.Then, through study and compared the several typical image processingmethod, designed peanut seeds image preprocessing method. Consist ofgray-scale image processing, image filtering, edge detection, image segmentation,morphological processing. We have studied the image feature extractionalgorithm, it can obtain the various characteristics of peanut seeds. Including area,perimeter, centroid, major and minor axes, equivalent diameter, roundness,elongation, compactness, color, etc. Select the four red peanuts and Lu peanutsthat have widely planted area and large quantity for feature extraction researchand analysis. Image preprocessing methods proved to meet the requirements offeature extraction. Finally, the use of LabVIEW and IMAQ Vision Development Kit Toolsdesigned the peanut seed feature extraction experiment platform. The paper haveanalyzed the feature extraction overall system architecture, detailing the mainfunction and each part of the user interface and design process diagram. Thisfeature extraction platform, feature extraction algorithm can be intuitive test datadisplay, easy interface to operate, easy to maintain and update. From theexperimental results obtained, this automatic identification system of peanutseeds that can accurately extract characteristic data of peanut seeds. It is a batchof peanut seeds sorting foundation.