Automatic Shooting Scoring System Based on Image Processing

Abstract

Today, many shooting venue that held shooting training used conventional methods to accumulate shooting scores. These conventional methods for shooting training will need more time and resources to do such a task. These paper describes an automatic shooting scoring system based on image processing for live shooting session. A camera is mounted in front of shooting target frame to capture every single shoot image. We use several image processing algorithms such as target ring detection, perspective transform, image subtraction, as well as morphological image processing. Contour detection method is used to perform a perspective transform, obtaining circle diameter and center circle position by using bounding box function by extracting detected contour and bullet hole position. Our experimental results, show that the accuracy of our method is 91%, based on the experiment by using a tiny circle sticker with a diameter of 7.62mm to simulate as a bullet hole image. We use 10 target sheets which there are 10 bullet hole images using circle sticker in each captured target sheet image.