Research On Wheat Spike Identification Based On Color Features

Wheat is one of the most important food crops in the world,acreage and production to occupy the first,using wheat as the main food of the population accounted for the world’s total population of 1/3,wheat production in the national economy occupies an important position,is foundation of the whole agricultural production.How to accurately and efficiently estimate the yield and wheat tillering rate,increase the unit area yield of wheat is an important research direction of wheat breeding expert.Automatic identification of wheat spike can not only used for yield prediction and tillering rate automatic calculation,and in the breeding selection,predictive density,spike rate detection also has important significance.Unit area grain number is an important parameter in the prediction of wheat yield,how to automatically identify the number of wheat from image is the key of automatic measurement.On The Henan Agricultural University networking platform,4 to 5 months wheat images are collected as research material,using a computer image processing technology as the main technical means,through the image pre-processing,the stalks and leaves,soil color feature difference image segmentation,through the form of learning processing value of wheat image,a series of corrosion,expansion of the operation,through refinement operation,extracting wheat skeleton using Harris corner detection algorithm,the realization of the computer automatic identification number of ears per unit area.And through the comparison with manual counting,verify the reliability and the accuracy of the recognition algorithm of wheat,ear recognition accuracy rate reached 95.5%.