Virtual Booth

Farm Watch is an innovative approach to enhance farm security and mitigate the issue of monkey disturbance in the farm. This project integrates Artificial Intelligence (Al), Internet of Things (loT) and renewable solar energy to create an intelligent scarecrow system. The system utilized Al algorithms to train a model for monkey facial detection. Once a monkey face is detected, the scarecrow is activated through ESP32, causing rotational movement of its hand and vertical movement of its head using two servo motors and a motor respectively. In addition, the system also triggers blinking of RGB LEDs in the scarecrow's eyes and emits loud sounds from a speaker. The architecture comprises two ESP32 microcontrollers, where one serves as a web server managing RGB lights and motor movements, and another functions as a client connecting to the access point ESP32 to control the servos. Simultaneously, a Raspberry Pi 3 interfaces with a trained model, camera and speaker The detection system provides the confidence level of the species identified to differentiate between humans and monkeys. To enhance system efficiency, the dual-core capability of ESP32 is utilized to perform tasks simultaneously using multitasking and variable sharing, Due to limited power supply in the farm, the Raspberry Pi 3 is powered by solar energy, while the ESP32 devices are powered by batteries. Moreover, the Wifi capabilities of ESP32 solved the problem of poor internet connectivity in the farm by establishing ESP32 as an access point. This proposed system ensures an effective and sustainable solution for farm security and monkey deterrence in an environment of limited power supply and poor intemet connection.