Recognition of bolt and nut using artificial neural network

This paper focuses on the recognition system of bolt and nut in real time for application in various industries, particularly the automotive industry. The objective of this study is to develop the image processing algorithm to get the normalized cropping images which would be suitable inputs for the learning process using Backpropagation Neural Network. Testing is done using a real-time visual recognition system. The Matlab software version 7.6 is used to integrate all algorithms, whereas the stepper motor differentiates the final result of bolt and nut in separate places. The result shows that the system can detect moving object accurately on the belt conveyor at a speed of 9 cm/sec. with an accuracy 92%.