Crop classification is gaining more attention in the medical and pharmaceutical fields as it benefits in producing medicines for managing life-threatening diseases. Medicinal Crop contributes prevent human and animal health by building medicinal properties on their roots, stem, and leaves. Further different parts of the Crop produces several concentration of distinct molecule compounds as a predominant component of medicinal usage. However manual analysis of the medicinal properties of the Crop is highly complex as it requires taxonomical skills. To alleviate those challenges, a hyperspectral imaging sensor has been employed to extract the biophysical traits of the Crops. Further classification of the biophysical properties is carried out using machine learning and deep learning model. In this article, an extensive study has been carried out on the analysis of the conventional architecture and its formulation to segment and extract the vital features of the Crop and classify it into suitable types for processing the hyperspectral images. Each architecture initializes the process with a dimensionality reduction process to extract only the significant information on transforming the original data into another feature space based on certain evaluation criteria. The reduced feature is analyzed on the classifier to discriminate the feature based on Crop classes on the objective function of the model. Particularly detection and identification of the Crop classes are effectively carried out on the spatial and temporal details of underlying land cover. However, Crop classification concerning spectral characteristics was obtained on the anatomical features and morphological features. Extracted features towards classification lead to several challenges such as large spatial and temporal variability and spectral signatures similarity between different objects. Extensive analysis of the architectures on results of the classes on various datasets is evaluated using performance measures.