Font Size: a A A

Key Technology Research Of Cucumbers Harvest Robot Based On Machine Vision

Posted on:2014-02-09Degree:MasterType:Thesis
Country:ChinaCandidate:L Z JinFull Text:PDF
GTID:2248330392460641Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
According to the latest statistics, the plant area of cucumbers inChina is more than1.1million hm2in2012, in nearly all of these farmfields or greenhouses, the only method to harvest cucumbers is pickingthem by human workers, which causes a great waste of the finite humanresource, in the background of acceleration of aging process. What’s more,the cost for human workers stands more than30%in the whole cost ofcucumbers production, so it will save a lot of human resource if wedevelop a harvest robot for cucumber production, and this will reduce thecost of cucumber production, increase the productivity too.Based on the research of the harvest robot done by researchershome and abroad, this project take the particularity of the cucumber intoconsideration. By studying and developing cucumber recognition andlocalization method, arms construction and control, and end-effectorconstruction and control, we make a systemic research on cucumberharvest robot. The main achievements of our research are listed below:1. The cucumber recognition and localization method based on RVM and I-Relief. When the camera gets an image, we need to recognize thecucumber from its background, and then we can localize it. Because thecolor of cucumber is very similar with the color of its background (e.g.leaves), it is hard to recognize it by using threshold segmentation of RGBlayers. Thus, we develop a prediction machine which is trained withknown pixel simples by using RVM and I-relief algorithm. This method ismore efficient than threshold segmentation or normal SVM method.Besides, we use a binocular camera to localize the position of cucumbers.2. Arms structure and control. After the recognition and localization systemget the position of the cucumber, robot system generates an order for armcontrol system, which will guide the arms of the robot carrying theend-effector construction move to a particular position.3. End-effector construction and control. After the arms finish its work,robot system generates a signal for the end-effector control system, whichfinally makes a catch and cut action done.
Keywords/Search Tags:Machine Vision, Cucumber Harvest, Robot
PDF Full Text Request
Related items