Today Elephant conflict is the most important problem. If we want to avoid this problem then it is very important for us to identify the elephant and understand the behavior that makes it easier to understand his moment. In such cases, individual elephant identification is one of the fundamental ingredients for the success of elephant conflict. Thus, in this paper, we shall explore and examine how image processing technologies can be utilized in analyzing and identifying individual elephant along with deep learning techniques. This system is mainly focused on the identification of individual elephant based on the color image of the elephant’s body. In our system, firstly we detect the elephant’s, Crop body reign, and then process a deep convolution neural network and the last one is to identify the elephant’s name and behavior. Then, we crop the elephant’s body region by using the predefined distance value. Finally, the cropped images are used as input data for training the deep convolution neural network for the self-collected elephant images are captured in the elephant rescue center at Tamor Pingla, Surguja, Distict, and Chhattisgarh, India. According to the experimental result, our system got an accuracy of 96.8% for automatic cropping in the elephant’s body region and 97.01% for elephants’ image identification. The result shows that our system can automatically recognize each individual elephant’s images. Our data to train the CNN learning algorithms. However, behavioral observation synchronized with relocations and acceleration records are often missing or unattainable in many elephants, hindering the application of supervised learning, making unsupervised learning a suitable tool for behavioral annotation of, movement paths in secretive or less studied elephants. Environmental and behavioral annotation of animal movement paths by machine learning improves understanding of the effects of environmental condition on animal movements and behavioral decisions.