1M.Tech, Mechanical Engineering, Universal Institutions of Engineering and Technology, Bollapur, Lalru
2Assistant Professor, Mechanical Engineering Department, Universal Institutions of Engineering and Technology, Bollapur, Lalru
Especially in the mechanical automobile industry where nuts and bolts are produced, a system that recognises the nuts and bolts has to be designed. The major objective of this work is to establish a method for recognising the difference in the separation of nuts and bolts. The data collection of 1000 pictures of nuts and bolts is collected for processing this system and the model is trained using the Convolution Neural Network. Many models are trained at different times with different batch size and have been documented with a proportional loss of validation. The trained template is stored and a picture is then supplied to the model to be classified into a nut or a bolt.In the classification of the nuts and bolts, our model was 95% correct. The language used was python for coding and the simulation was done with anaconda software.
Dataset, Open CV, Convolutional Neural Network, Epoch, Batch Size
