Lichens are interesting composite organisms that evolved and diversified after a symbiotic association between algae and fungi and lichens are estimated to cover roughly 10% of terrestrial ecosystems. The limited number of images in datasets makes it difficult to classify the lichen classes and the dependability rate of the existing study is still quite low. In this paper, a novel deep-learning model is proposed for the classification of lichens using aerial images. The input aerial images are gathered from western ghats and the collected images are pre-processed utilizing a Contract stretching adaptive histogram equalization (CSAHE) filter to increase the image quality. The Mask RCNN model is implemented to extract the relevant features from the images and also segment the region of the enhanced images. The Deep neural network is used for classifying the lichens from the western ghats. According to the result, the proposed model attains a 99.12% success rate for the classification of lichens. The proposed Mask RCNN enhances overall accuracy of 2.53%, 6.39%, and 1.88%, better than RNN, CNN, and RCNN. The proposed DNN improves its reliability by 8.13%, 4.45%, and 0.87% better than FNN, GNN, and DBN respectively. The proposed model enhances the overall accuracy of 38.33%, 10.12%, and 4.12% better than DCNN, CNN, and XGBOOST.