Pet cat behavior recognition based on YOLO model

With the progress of the times and the rapid development of science and technology, machine learning and artificial intelligence are increasingly used in transportation, logistics, and homes. In terms of pets, pet monitoring has also become very popular in recent years. Therefore, this study for the real-time monitoring of home pets, using the raspberry pie as a monitoring system, proposed a raspberry pi-based YOLOv3-Tiny identification system YOLOv3-Tiny method with rapid detection and better boundary frame prediction. One thousand one hundred twenty-eight pictures of cats' movements were collected in the room for marking and training. Finally, according to the input image categories, the results of six cat action categories were output. They were sleeping, eating, sitting down, walking, going to the toilet, and search on a trash can. The average accuracy was 98%. Through image recognition, the images were sent to the user's mobile phone app and computer. When the cat goes to the toilet for too long or flips through the trash can, the system will instantly send a message to the owner's mobile phone to achieve an instant preventive remote pet monitoring system.