Efficient irrigation management in agricultural and urban landscapes is important for water conservation, crop productivity, and plant health. The incorporation of IoT in irrigation systems, with smart sensors that monitor soil and climate conditions, allows for precise decision-making and optimized water usage. The main objective of this research is to implement a smart irrigation system as an efficient watering tool for turf surfaces, providing area recognition and generating appropriate irrigation using computer vision. The detection algorithm is based on YOLOv5 training to recognize turf and utilizes Python to detect contours of grass and objects. The system incorporates electronic components to achieve efficient and automated irrigation. For data acquisition, the FC-28 soil moisture sensor, and the ESP32-CAM as an IP camera were chosen. Monitoring is conducted via a mobile application developed in Android Studio for user compatibility, culminating in an integrated system that demonstrates the practical application of computer vision in smart garden irrigation. The results obtained showed a 95.6% accuracy in grass detection. Tests conducted in a garden indicated that the smart system can be beneficial for efficient irrigation by accurately identifying contours of objects and green areas.