Quality Analysis of Steel Rolling Bar by Comparing ANN, CNN and SVM Algorithm
Quality testing is one of the important step in all major industries. In this project we are concentrating on steel rolling mills which manufacture steel rods for building purpose. Therefore quality of such rods becomes an important cause. There are various defects detected in steel rods. Few of the defects we are going to discuss about are surface defect and deformation of rod. Therefore we capture the image while the factory is on working. This is done by using Raspberry pi and pi camera and transferring the image through local Wi-Fi network. Compare the image captured and detect the defected rods whether it’s a surface defect or deformation or is it good quality rod. In this proposed project we try to compare three types of algorithm and conclude which algorithm is best suited for indicating whether the captured image is defected or is it a good quality rod. The three algorithm that are compared in this paper are Artificial Neural Network, Convolution Neural Network and Support Vector Machine. Keywords - Artificial Neural Network (ANN), Convolution Neural Network(CNN), Raspberry pi, Steel rods, Support Vector Machine(SVM).