This repository stores code used for a human detection project (part of the Autonomous Systems Practical Extension at Rijksuniversiteit Groningen). In the project we used both laser and camera modules of a robot in order to distinguish humans from non-humans. The summary of the process follows. First data was derived from features of laser point clusters which were then modelled with the help of adaptive boosting to detect areas of interests. Then a support vector machine was used to classify the histogram of gradients of the detected area as either containing a human or not.