Animal attacks are a frequent occurrence, with elephants being prominently involved in these incidents. Elephants, in search of food and water, enter human establishments and crop fields, which will eventually lead to conflicts. Conflicts sometimes end up resulting in the loss of lives on both sides. To prevent these conflicts from occurring, this paper proposes a real-time elephant detection system based on the Regional Convolution Neural Network (RCNN), which is embedded into the Raspberry Pi. If an elephant or an elephant herd is detected in the designated area, an alarm and a notification message will be given to the concerned people. The algorithm has also been compared with other object detection algorithms, such as You Only Look Once (YOLO) and Single Shot Multibox Detector (SSD).