Detekce průstřelů v terčích

Abstract

This diploma thesis deals with design and implementation of mobile application with Android operating system, whose task is to detect bullets in International pistol targets 50/20 using camera on phone. The solution is based on image analysis methods and convolutional neural networks. Android application will automatically score the shooter based on detected bullet holes, transformation of the target into normalized shape and using a template with information about the distribution of point values in the target.