| With the higher demand for speed and precision in industrial production,machine vision technology which has advantages of non-contact,fast speed,high accuracy and strong anti-interference ability has been widely introduced into industrial systems.Object localization and identification is an important part of machine vision research,and template matching is often used to identify and locate workpieces because of its simplicity and effectiveness.In this paper,a fast template matching algorithm based on transformation of edge points set is proposed.The main contents are as follows:Firstly,a machine vision system which selects a backlight illumination is built after analyzing the requirement of workpiece localization and identification.Secondly,an edge tracking algorithm is proposed to record the coordinates of edge points,which finds the next tracking point in the way of calculating the deviation parameters of each point in the neighborhood of the current point by marking the inner edge points.Thirdly,the position and degree of each object in image is described by centroid,principal axis and minimum enclosing rectangle and the real position and degree can be obtained by the transformation matrix of the image coordinates and the world coordinates which is calculated by camera calibration.Fourthly,a multi-objective template matching method based on transformation of edge points set is proposed.The transformation parameters are obtained by marking the pairs of feature points.The similarity between template and target points set can be described with Hausdorff distance and then the object can be identified.Lastly,the system software of workpiece localization and identification is developed and the speed and accuracy of the system are tested by experiments.The results show that the system can meet the location accuracy and real-time requirement of industrial measurement. |
