Plant Recognition System based on Leaf Images:

Abstract

Abstract: Plant plays a vital role in the environment. Nature has enormous members of plants, identifying them and classifying them is an important task for botanists. They are still finding difficulties in recognizing those plants and it is complex, time-consuming, and due to the use of specific botanical terms, frustrating for the non-experts. There are various ways to recognize a plant, like a flower, root, leaf, fruit, etc. But usually, plants are recognized by leaf and their characteristics like shape, texture, vein structure, color, etc. The availability of relevant technologies, such as digital cameras, new techniques in image processing and pattern recognition thoughts, leads to researching plant leaf recognition systems using image processing techniques rather than using the other parts of the plants. Searching of research papers related to the domain was done in the IEEE computer society digital library, Springer Link, Science Direct, ACM Digital Library, Academia, and other research sites using the search terms like plant leaf identification, plant leaf prediction, plant leaf classification, plant leaf recognition, and image processing techniques. Then started to work on the Systematic Literature Review (SLR) with forty research papers; out of forty, most articles are related and some are not related. Then these articles are filtered and sorted to SLR. Nineteen papers which are published in the past ten years and mostly related to my topic and methodology, were selected to perform SLR. Then the methodologies of the studies were analyzed to identify different preprocessing and feature extraction methods that researches have employed to identify leaves. Then classification accuracy was compared with related papers and traced the optimal range of the accuracy for leaf recognition, which will be the benching accuracy level for the suggested study.