Plant Leaf Recognition based on Mask R-CNN

Plants, as the main form of life and an important component of human life, are often as the preferred objects for automatic plant recognition research due to their distinct features, prominent internal textures, and significant differences in appearance. Firstly, a plant leaves recognition dataset containing 805 images was constructed. These leaves are classified as 8 different plant species, including blueberry, apple, cherry, grape, strawberry, capsicum, peach and potato. Subsequently, Mask R-CNN was used as a basic method to provide a baseline leaves recognition result. The plant leaves dataset has been made public on Baidu Network Disk and the link is https://pan.baidu.com/s/1YIE-QO25lQIlvQf5QFbKjQ?pwd=mn89.