Design, development, and performance evaluation of a GPS and machine vision-based navigation system for an intelligent field guard robot

An innovative approach to protecting farmland from wild animals is the use of an intelligent field‑guard robot. In this study, a fully featured field‑guard robot was designed and developed using image processing and a global positioning system (GPS). To achieve this, the robot’s tracking system, which combines a machine‑vision camera and a GPS receiver, was evaluated. For automatic tracking, several control points were established within an experimental field, and their latitude and longitude coordinates were provided to the robot to enable GPS‑based navigation. As the robot moved through the field, it captured and processed images to detect and pursue moving objects, such as approaching or attacking animals. The performance of the tracking system, the image‑processing algorithm, and the robot’s ability to detect and chase animals were investigated. The results showed that the robot’s tracking system performed better in sunny weather compared to cloudy and partly cloudy conditions. To evaluate the image‑processing algorithm, RGB, HSV, and Lab color models were examined, and the RGB color model was found to be the most suitable. A normalized standard deviation of ۱۰% in the image provided the best performance for detecting the attacking animals. Evaluation of the robot’s performance in repelling attacking animals showed promising results, with successful repulsion of four out of five attacking animals under sunny and partly cloudy weather conditions.

کلیدواژه ها:

Attacking animals ، Camera ، Control points ، GPS ، Image processing