This document discusses using machine learning to recognize animal species in images taken by motion-triggered camera traps. It explores using convolutional neural networks (CNNs) to classify images and combining image recognition data with metadata like temperature, date and time. Two CNN models (Caffe and TensorFlow) were tested on labeled camera trap images, with retraining improving accuracy from initially inaccurate results. Optical character recognition was also used to recognize temperatures in images and combining data sources could help analyze animal species trends related to time and temperature.