Object Detection Approach Using YOLOv5 For Plant Species Identification

In the modern era of agriculture and horticulture, biodiversity conservation requires plant species identification skills, and automatic detection is a challenging and interesting task. However, many factors often make some people mistaken in recognizing plant species that have unique and varied visual characteristics, making manual identification difficult. This problem requires an effective and accurate model for identifying plant species. So this research aims to produce a model to identify plant species that are effective and have a high level of accuracy. This research offers the use of the YOLOv5 algorithm method. The training process with epoch 200 and 53 minutes with a total of 1,220 images. Based on the results of the model performance test, the mAP value was 85.73%, precision 98.27%, and recall 94.36%. During testing, the model can identify plant species accurately on single objects and multiple objects. The results of this research show that the proposed method is successful in identifying plant species accurately.