tOWARDS AUTONOMY PERCEPTION LIBRARY (tapl)

The goal of this library is to provide an open-source platform for perception algorithm pipelines. It provides methods for several computer vision algorithms such as Structure from Motion (SfM), Camera Calibration from a single image, Visual Odometry, Bundle Adjustment, Feature Detection and Tracking, Panoramic Image Stitching, etc. TAPL also provides methods for point-based processing such as KD tree implementation, Euclidian Clustering, Plane Fitting, and so on.

Quite often, we find ourselves in situations where we quickly need to implement certain perception algorithms as a part of a larger goal. Driven by curiosity, we also desire to learn how one would implement them from the ground up and even improve them for better performance and faster run-time. Open-source code implementation allows us to do just that and is the goal of TAPL. We provide the implementation of computer vision algorithms that can be used as plug-and-play modules and are easy to play around with for various use-cases and even improve upon.