Image Processing for Drones Detection

摘要

Presently, Drones Quadcopter has caused tremendous problems threatening the military security boundary and the area under Thai Army's surveillance. This article presents a conceptual framework for detecting and tracking unmanned aircraft by applying image processing. Study and analysis Compare advantages and disadvantages in order to develop and provide suitable equipment and tools Replacement of imported high-priced With limiting factors, namely distance, time, terrain And interference resistance The researchers used the camera USB 3.0 Cameras and Open Source Computer Vision (OpenCV) as a software development library on Linux operating systems to automatically record movies. In order to bring the image to analyze and distinguish the object classification by using Machine Learning through a sample of correct information and some incorrect information. Which when detecting, the installed device will calculate the coordinates detected, lock the target, track the movement Voice notification and report. The experiment was conducted by using Anova to test various factors affecting the detection, such as speed, light, color and size in 350 feet with the equipment we installed. The results of the experiment concluded that there was only a speed that had an effect at 22.65.