ThermalTracker v2 is software for remote sensing of moving animals and other objects in stereo thermal video. The inventive features are: Three-dimensional tracking of moving animals and other objects in near real-time using a novel stereo vision algorithm. Automated identification of objects based on features of the motion track and object size estimates. The ThermalTracker v2 software processes thermal video from a stereo pair of thermal cameras and extracts the 3D flight tracks of birds and bats and other flying objects. The system uses a novel approach to generating the 3D tracks that does not involve conventional tracking methods, is computationally efficient and can run in realtime. The method works by first extracting the 2D flight tracks from each camera independently, and then applying stereo vision processing to the resulting pair of flight track images. The flight track images are generated by collapsing a sequence of video frames into a single image that shows an entire flight track through the camera's field of view. The method makes use of both spatial and temporal information in order to align and match the tracks in the two images. The temporal information helps overcome the challenge of matching objects in a stereo pair of thermal images, where the shapes and spatial features in the images may be somewhat blurry due to motion and the low resolution of the thermal cameras relative to visible spectrum cameras. The ThermalTracker v2 software produces a compact, visual representation of the observed object motion along with a detailed characterization of each motion track.