LBEF RESEARCH JOURNAL OF SCIENCE, TECHNOLOGY AND MANAGEMENT

Author(s):Uday Kumar Sah, Jyotir Moy Chatterjee

Abstract:Deep learning is quite popular in the field of image processing. consequently, it is mostly used for identification and recognizing different subject and object in the images. Image recognition is the most innovative and revolutionary field of artificial intelligence which deals with pixels pattern to map their object by using pooling or edge detection and sketching methods. In this filed one of the most problem that is commonly facing by any botanist. Because they did not recognize any plant species because lack of proper identification of each and every species of plant and it is quite vast diversity in the context of Nepal. In this work, we were finding out how we going to resolve plant identification and detection problem by using the power of convolutional neural network (CNN) by using Plant@Nets Datasets 2 were 3200 images with 53 different classes. We will also find out what are the accurate and suitable steps to develop PlantRecog Application by using Convolutional neural network. Secondly, the researcher would like to include the process and mechanism of PlantRecog application development process and how the process 50 x 50 greyscale images to identify any plant species by using Convolutional neural network (CNN). This problem is directly belonging to the classification of Machine Learning (ML). we also implemented the best suitable waterfall software development methodology to carries out the complete work of PlantRecog applications. we give complete information about which kinds of hardware and software required for the development and deployment of PlantRecog Applications. we used primary and most popular libraries such as TensorFlow, Keras, Scikit learn and pillow to process the image data and allows algorithms to learn from them. The overall process of development carries out by using the Python programming language.

Keywords:Image Recognition, Plant Identification, Machine learning (ML), Deep Learning (DL), Neural Networks (NN), Convolutional neural network (CNN), Greyscale