Section: New Software and Platforms

Perception Tools

Participants : David Filliat [correspondant] , Louis-Charles Caron, Alexander Gepperth.

Of 3-D point cloud

Participants : Louis-Charles Caron [correspondant] , Alexander Gepperth, David Filliat.

This software scans the 3-D point cloud of a scene to find objects and match them against a database of known objects. The process consists in 3 stages. The segmentation step finds the objects in the point cloud, the feature extraction computes discriminating properties to be used in the classification stage for object recognition.

The segmentation is based on simple assumptions about the geometry of an indoor scene and the movement of a wheeled mobile robot. The floor plane coefficients are known a priori and are used to eliminate from the point cloud all points that are close to this plane and have a normal perpendicular to it. The floor plane coefficients also allow the detection of walls. Successive RANSACs are run to find planes that are perpendicular to the floor plane, and contain a large number of points. With these large structural regions removed, the only points remaining in the point cloud are the objects in the scene. These objects are separated by clustering the points based on a distance criteria. Close-by points are considered to form a single object.

Objects are characterized by their shape, texture. The texture information is encoded as a histogram that apprioximates the form of the distribution of color values in the object. A separate histogram is built for the red, green and blue channels. The shape of an object is encoded by computing thousands of randomly chosen Surflet-pair relation features and compiing them into a histrogram of occurrence.

The classification is done by a 3-layer feed-forward neural network. The network is trained on a dataset of point clouds of 53 objects. After training, the neural network is run on the features computed from each object detected in the segmentation stage [86] .

PEDDETECT: GPU-accelerated person detection demo

Participant : Alexander Gepperth [correspondant] .

PEDDETECT implements real-time person detection in indoor or outdoor environments. It can grab image data directly from one or several USB cameras, as well as from pre-recorded video streams. It detects mulitple persons in 800x600 color images at frame rates of >15Hz, depending on available GPU power. In addition, it also classifies the pose of detected persons in one of the four categories "seen from the front", "seen from the back", "facing left" and "facing right". The software makes use of advanced feature computation and nonlinear SVM techniques which are accelerated using the CUDA interface to GPU programming to achieve high frame rates. It was developed in the context of an ongoing collaboration with Honda Research Institute USA, Inc.

A Python OptiTrack client

Participant : Pierre Rouanet [correspondant] .

This python library allows you to connect to an OptiTrack from NaturalPoint (http://www.naturalpoint.com/optitrack/ ). This camera permits the tracking of 3D markers efficiently and robustly. With this library, you can connect to the Motive software used by the OptiTrack and retrieve the 3D position and orientation of all your tracked markers directly from python.