Boominathan. Perumal’s research while affiliated with Vellore Institute of Technology University and other places

This study introduces a method for fruit counting in agricultural settings using video capture and the YOLOv7 object detection model. By splitting captured videos into frames and strategically selecting representative frames, the approach aims to accurately estimate fruit counts while minimizing the risk of double counting. YOLOv7, known for its efficiency and accuracy in object detection, is employed to analyze selected frames and detect fruits on trees. Demonstrated the method's effectiveness through its ability to provide farmers with precise yield estimations, optimize resource management, and facilitate early detection of orchard issues such as pest infestations or nutrient deficiencies. This technological integration reduces labor costs and supports sustainable agricultural practices by improving productivity and decision-making capabilities. The scalability of the approach makes it suitable for diverse orchard sizes and types, offering a promising tool for enhancing agricultural efficiency and profitability. The researcher compared YOLOv5n, YOLOv5s, YOLOv7, and YOLOv7-tiny with eight-sided imaging techniques around the tree. The experimental results of YOLOv7 with the eight-sided technique performed best and achieved a count accuracy of 97.7% on a single tree in just 17.112 ms of average inference time. On multiple trees, it is 95.48% in just 17 ms of average inference time, with the help of an eight- sided method on tree images.

Getting the size of any fruit on a tree is not an easy task especially mango fruit, because of its irregular shape, it is not easy to model with its shape. To do so we need the size of the fruit in length and width. In this journal, the researcher used the aruco marker for size estimation in computer vision for size recognition of the fruit, in image processing concepts, and got greater accuracy of the fruit size in real-time with good accuracy using an image processing and deep learning algorithm at detection. Objective: Horticulture farmers need to do some extra activities to get better yield like trying to know fruit shape, and fruit size at the time of maturity or before plucking fruits from the tree which will help farmers to get as per their predicted price while selling the fruits to the market nowadays. But the farmers are selling their fruits without knowing the size and shape of the fruit and their hard work because there is no measuring device to measure the farmer’s hard work, but there is a possibility to measure the size of the fruit which is a major drawback to them. To overcome this problem, the researchers tried to find a better solution for the farmers. Methods: Researchers applied a deep learning model named YOLOv7, Semantic Segmentation to get fruit size using an aruco marker. The researchers proposed a technique to help farmers that detect markers and the fruits of images and predict the size of the fruit at multi-targets. For this work, a custom dataset was created by collecting mango fruit frames from on-tree-mango-360° recorded video and the researcher did not augment the dataset. After training and validating this model, the performance was tested on the test dataset. Results: The contributions of this article are: The researcher developed a procedure to get the mango size from an image. The researcher implemented and tested a model to detect Banganapalle mango fruit in different challenging situations using YOLOv7 with Semantic Segmentation. Finally, the model achieved very good results on fruit size estimation. The training and testing results of YOLOv7-SS-AM show that the aruco marker-based model is superior to the manual size prediction, with good accuracy too.