Tree species classification is an active area of research, in forestry related tasks, which is quite challenging but has useful applications. It is difficult to meet the challenge due to complexity of the problem as well as due to non-availability of datasets for research purposes. In this thesis, we are focusing on creating a dataset for classification of tree species that exist in Pakistan and subsequently use it for the species classification. It is a novel piece of work for the species that exist in the homeland. We are devising deep learning-based algorithms for the classification. Deep learning has been phenomenal in recent years in image-based classification, and we are employing state of the art convolution neural network (CNN) based methods for the above-mentioned tasks. There are further challenges, i.e., scarcity of tree species images for training the network, which will cause overfitting. In order to meet with the challenge of scarcity of images and subsequently the problem of overfitting, we are employing data augmentation to mitigate the effects. Upon completion, this work will be very useful in forestry applications.