Comparative Analysis of Image Processing Object Detection Models: SSD MobileNet and YOLO for Guava Application

Object detection has emerged as a crucial task in computer vision and finds its application in various domains. The present study aims to compare two state-of-the-art object detection models, SSD MobileNet and YOLO, for detecting guava by image processing. To accomplish the study's objective, a dataset of guava fruit images and videos was collected and used to train the models. Both models were evaluated on various performance metrics, including accuracy, speed, and precision. The results on COCO dataset showed that YOLO outperformed SSD MobileNet in terms of both accuracy and speed. YOLO achieved an average precision of 0.93, while SSD MobileNet achieved an average precisely of 0.86. Furthermore, YOLO could process images and videos faster than SSD MobileNet, with an average processing time of 0.2 s per image, compared to 0.3 s for SSD MobileNet.