Quality Control Using U-Net: Detecting Defects In Leather

Recently, there has been a significant amount of attention in the literature towards computer vision algorithms, particularly those that focus on semantic segmentation applications. This is largely due to the availability of data to train models, as well as the accuracy of these algorithms. Among the various computer vision applications, the U-Net has gained popularity due to its reliable accuracy, simplicity in construction, and ease of application. Despite the advantages of this network structure, there are still some unclear aspects within the U-Net that have not been sufficiently covered in the literature — to the best of our knowledge —. This study seeks to clarify and explain these ambiguous points, using different architectures and on the MVTec dataset to demonstrate our proposed setups and their efficacy.