Identification of Medicinal Plants by Visual Characteristics of Leaves

Abstract

A vital ability for foresters and botanists is the ability to identify plant species from photos of leaves, bark, and needles. Plants are identified by human professionals using scent and visual traits. If therapeutic plants are incorrectly identified, there might be detrimental effects. Using visible morphological characteristics, such as leaf shape, color, and texture, plant identification may be done automatically. This paper demonstrates how image processing and machine learning techniques were used to accurately find rare medicinal plants. A unique feature combination is used to classify the leaves. Validation rates of up to 95% were obtained while testing more than 20 plants.