Is Convolutional Network Competitive for Vision Method in the Furniture Dowel Quality Control?

摘要

The paper presents a method of the furniture dowel quality control based on the convolutional neural network. We applied AlexNet, testing the ability of transfer learning. The paper presents details of the method and experiments performed to justify our solution on images datasets collected from real dowels production. The outcomes are compared with the results of baseline vision method. The experiments show that both methods have their strengths and weaknesses, but they complement each other.