The Faculty of Applied Engineering has a new doctor! Dr. Izaak Van Crombrugge defended his doctoral thesis on the 2nd of December, with professor Steve Vanlanduit and professor Rudi Penne as promotors. His doctoral thesis is titled ‘Safety Tracking with Range Camera Fusion’.
In this doctoral research, dr. Van Crombrugge investigated how to improve worker safety using depth cameras. These are cameras that measure the depth for each image pixel. To avoid accidents at work, employers can track the position of all personnel in the room and adjust the behavior of the machines.For example, a crane dropping a heavy container on personnel standing underneath can be prevented. By using depth cameras, the tracking is very reliable. In addition the system is completely anonymous so the privacy of the workers is not being invaded.
In his research, dr. Van Crombrugge also takes into account the posture of the admitted workers. From the camera images, a simple 3D skeleton can be detected. The camera then automatically calculates the ergonomic score for this posture based on the angles of the joints. In this way ergonomic problems can be detected easily.
To correctly align the images of different cameras, extrinsic calibration is needed. In his doctoral research, dr. Van Crombrugge provided two convenient calibration methods to find the pose of a set of cameras. No overlap is needed between the different fields of view, allowing for efficient camera placement.
Dr. Van Crombrugge developed a robust method for people tracking using range cameras: Density Map Tracking with Blob Splitting (DMT-BS). Multiple cameras can easily be added to resolve occlusions and to enlarge the observed area. The method can be used to track any moving object in an otherwise static environment, as the detection does not rely on a specific human model.
Its strength lies in its simplicity, making the behavior predictable and opening possibilities to be implemented on low-cost hardware. From the point cloud delivered by the depth sensor, a 2D density map is formed in floor coordinates followed by basic 2D tracking. Robustness is enhanced using a simple but effective blob splitting technique. Tests show that the camera position, depth noise, and extrinsic calibration errors have little influence on the tracker’s performance. The proposed method was tested on three depth tracking datasets, reaching significantly better MOTA (Multiple Object Tracking Accuracy) scores when compared to two state-of-the-art depth-based trackers.