Precision Fruit Harvesting: Stem Detection and Cutting Techniques using Image Processing for Robotic Systems

To enhance robotic fruit harvesting systems, the proposed methodology uses image processing techniques to recognize orange stems. The method highlights only the stem of an orange fruit in an image by utilizing gray-scale conversion, contour-based filtering, Gaussian blur, and Canny edge recognition. The technique can be incorporated into agricultural robots to improve fruit harvesting accuracy. After detecting the fruit stem, it is crucial to ensure precise cutting of the fruit to prevent damage to both the fruit and the tree. The paper also assesses the performance of the method in terms of robustness in complicated situations and detection accuracy, demonstrating its potential for real-world agricultural applications.