A Cow Crossing Detection Alert System

Abstract

Artificial intelligence is rapidly growing in recent years and has derived several branches of studies such as object detection and sound recognition. Object detection is a computer vision technique that allows the identification and location of objects in an image or video. On the other hand, proper recognition is the ability of a machine or program to receive and interpret dictation or to understand and carry out direct commands. This paper presents a cow crossing detection alert system with object detection and sound recognition capabilities. The proposed system aims to protect the driver from animal-vehicle collision. A data-driven deep learning approach is used for cow detection. Consequently, the cow detection module is integrated with a Raspberry Pi device to perform real time monitoring. The proposed cow crossing detection alert system will send an alert message to relevant units like the road transport department to take further actions if a potential animal-vehicle collision is detected on the road. Experimental results show that the proposed cow detection approach yields a mean average precision 0.5 (mAP 0.5) of 99% in object detection and 100% accuracy in sound recognition, demonstrating the system's practical feasibility.

Keywords

• cow crossing detection
• iot
• object detection
• sound recognition
• yolo