文章摘要
BP neural network prediction of rape maturitybased on pod hue H value
  
DOI:
Author NameAffiliation
宗望远,王佳伟,王淑君,赵颖彪 华中农业大学工学院湖北 武汉 430070 
Hits: 1376
Download times: 533
Abstract:
      In order to solve the problem that the proper time of harvesting is difficult to determine in the current process of rape combine harvester,this paper carried out prediction studies of digital coloring in rape pod RGB(red,green and blue)and the grade of maturity. Through manipulation RGB value in effective pixel points that were collected from picture,thus the hue H value of rape pod after 1-14 days of rape emergence was obtained. Taking the pod hue H value of 1-12 days after rape emergence as training data set,taking the pod hue H value of 13,14 days after rape emergence as validation data set,using MATLAB software,the mathematical model between the rape pod hue H value and the number of growth days after 204 days of rape emergence was established,taking use of the pod hue H value,the judgment of rape pod maturity was achieved. The rape maturity model was established based on BP neural network and the equation result was verified through the later test. Results showed that rape pod hue H value increased with the day increasing. The predictive rape pod epidermal H value in the 13th,14th day were 64.4640,64.4723,while 64.4296,64.4307 respectively as the actual value,with the difference of 0.053% and 0.065%. The test predictive value was consistent with the actual value of measurement result. Rape maturity was delineated according to the rape pod hue H value. Green mature rape pod hue H value was 64.00-64.19,yellow rape pod hue H value was 64.20-64.39,ripe rape pod hue H value was 64.40-64.49,as the classification criterion of rape maturity.
View Full Text   View/Add Comment  Download reader