STUDY OF AUGMENTATION METHODS IN THE PROBLEM OF STONE SEGMENTATION ON A CONVEYOR BELT OF A MINING ENTERPRISE

Kalashnikov V. A.

This paper examines various image augmentation methods in the problem of segmenting stones on a conveyor belt of a mining industry. The use of the deep learning algorithm MaskRCNN and the PyTorch library to solve the problem is considered. Using the IoU metric as an example, the quality of augmentation under various transformations is analyzed. The most and least effective approaches were identified and their qualitative analysis was carried out.

Keywords:mining, quality control, artificial intelligence, neural networks, deep learning, augmentation