Image Segmentation and Classification: Alien Plant Detection Using MATLAB

Abstract

Biological diversity is threatened by invasive alien plant species (IAPs) that have spread outside of their natural range. They are not natural to the ecosystem and may cause economic or ecological harm. They have a detrimental influence on biodiversity, resulting in the decrease or extinction of native species because of water competition and disturbance of local ecosystems and ecological processes. IAPs have harmed natural biodiversity in nearly every ecosystem type on the planet and are one of the most serious threats to biodiversity. Traditional methods used to identify and detect invasive alien plants such as utilizing direct sampling field-based techniques or making visual estimation have provided average success. These methods are prone to errors, time-consuming, and labor-intensive. Remote sensing techniques offer a concise swift approach for detecting and mapping invasive alien plants. However, remote sensing hardware technology such as light detection and ranging (LIDAR) is expensive. The emergence of computer vision and machine learning has provided an inexpensive alternative that can be deployed from a mobile device. This paper illustrates the application of computer vision in the detection of invasive alien plants in South Africa.