23-10-2010, 09:33 AM
The design and implementation of an intelligent machine vision and sorting system is described in this article. It can be applied to sort objects in an industrial environment. The sorting systems generally work based on the geometry driven the textural components of the object. The textural analysis of pixel content is used as the methodology in this project. An an artificial neural network is used to perform the recognition task. The methods like the fuzzy logic and support vector machines were also available but the neural network was preferred because of its relative simplicity. The communication between the main computer running the intelligent recognition system and the remote robot control computer located in a plant environment is implemented through bluetooth. Image of the objects is taken and compressed before using them for the feature vectors extraction by using principal component analysis. This compressed data transmitted via the Bluetooth channel to the remote control computer for recognition by the neural network. It also performs the task of recognition and the robot control computer is sent a control signal. The computer then guides the robot arm to place the object in an allocated position. The performance analysis of the proposed intelligent vision and sorting system is done under different conditions and the favourable points in the outcome are relatively immune to noise, capacity to generalize , fault tolerance etc.
Get the report here:
http://ir.dut.ac.za/handle/10321/494
Get the report here:
http://ir.dut.ac.za/handle/10321/494