DEFECT GOODS INSPECTION SYSTEM USING CONVOLUTIONAL NEURAL NETWORK WITH TELEGRAM BOT APPLICATION

Abstract

All this time, in detecting the defect in the final product in the manufacturing industry, especially the head of the dolls, was conducted manually by the inspector or human dependencies. This research tries to introduce the use of Convolution Neural Network (CNN) in the manufacturing industry to detect defects in the final product. CNN is made up of neurons with weight, bias, and activation functions. When CNN applies a convolution (filter) of a specific size to an image, the computer obtains new representative information as a result of multiplying that portion of the image with the filter. To solve the problem mentioned above, Python programming is used to program the CNN process and Telegram Bot for the user interface, and lastly, the camera is used to capture the image of the final goods. The test result shows that the system is performing as expected with 90% accuracy.