Real-time Plant species recognition under unconstrained environments is a challenging and time-consuming process. The objective of this research work is to create a large real-time dataset of plant images and to develop a system for efficient plant species recognition. A self-developed dataset named ‘Leaf-12’ is formed using the Indian plant species. The dataset includes twelve plant species. The images in the dataset are captured by varying the illumination, scale changes, orientation or viewpoint modification, and different backgrounds. Four approaches are tested for real-time plant species recognition. They include the Conventional image processing method, Neural Network, Single Deep Learning (Convolutional Neural Network (CNN)) Architectures, and Dual Deep Learning Architectures (DDLA). The methods are evaluated using four datasets, namely, Flavia, Folio, Swedish leaf, and self-developed Leaf-12. DDLA architecture outperformed other proposed methodologies for plant species recognition.
Dr. V. Sathiesh Kumar
Assistant Professor,
Department of Electronics Engineering,
Madras Institute of Technology Campus,
Anna University,
Chromepet,
Chennai-600044.
For Research Poster click here...
For record slowed gif click here...
For more details click here...