In order to solve the problems of intensive fruit planting scale, rising labor costs, and rural labor force loss, a fruit-picking robot based on intelligent Internet of Things is designed in this paper. Fruit-picking robot uses raspberry pie 4B and STM32 board as the main part. The employed modules include USB high-performance camera, ultrasonic ranging, color sensor, wireless transmission, gear control, GPS, laser radar ranging and gyroscope. Using YOLO algorithm and D-H parameter method to achieve accurate identification of target fruit, accurate picking of ripe fruit and stable mobility of the machine. This can improve the adaptability to environmental changes for the fruit-picking robot. Users can remotely control the fruit-picking robot through the cloud, which can improve the production efficiency and quality of fruit agricultural products, and support modernization and sustainable development of agricultural production.