Exploring the Capability of NASNet Model for Plant Classification on GRASP-125 Dataset

Plant classification and identification, which has strong ties to both computer science and taxonomy, is becoming an increasingly dynamic area of research within the field of computer recognition. Researchers have utilized different techniques including image processing, pattern recognition, and machine learning to develop systems capable of accurately identifying and classifying plants. Access to large image databases and advances in image processing and artificial intelligence have led to the development of highly accurate plant identification systems. These systems have the potential to greatly reduce the amount of time and proficiency required for plant identification, which are only achievable by skilled taxonomists. In this study, we experimented with the Greek vAScular Plants (GRASP) dataset, consisting of images across 125 different plant species native to the mountains belonging Oiti and Parnassus in central Greece, to perform automatic plant identification using a deep learning approach namely NASNet. These plant species, which include both rare and aesthetically pleasing varieties, are often found along popular hiking routes in the region. Our method for plant classification utilizes deep learning techniques on the GRASP-125 dataset, with the goal of improving the accuracy and efficiency of plant identification.