Jörg Velten

A multidimensional wave digital filter bank for video-based motion analysis

Multidimensional wave digital filters (MDWDF) exhibit the same desirable properties as 1D WDFs, most notably including passivity and therefore guaranteed stability as well as high robustness. A possible application for such MDWDFs may be found in motion analysis of image sequences by means of filters with fan-shaped transfer functions, where content with specific movement information can be extracted. For that matter, a parallel filter bank is needed to differentiate object motion into separate classes. In this paper, a new specialized MDWDF fan filter structure is introduced, possessing both reduced computational complexity and memory requirements compared to existing approaches. Additionally, part of the processing can be shared among all bands, further increasing efficiency.

Multisensor data fusion for advanced driver assistance systems - the Active Safety Car project

Driver assistance systems support overstrained and affected drivers and become more and more essential for series-production vehicles. Object detection and segmentation is one of the most challenging research topics in this field. In order to warn the driver or automatically break before a potential collision, objects intersecting the path of the host vehicle have to be detected and classified. Most recently developed approaches are based on two dimensional image processing, sometimes in combination with a tracking algorithm associating detections in consecutive frames to one and the same object. Further robustness is achieved by multisensor data fusion, i.e. information by two or more different sensors (e.g. camera and radar data) are fused in order to get a much more reliable result. Another aspect for safety applications is communication between cars, which provides additional sensor locations and thus also requires data fusion technology. Two different approaches for data fusion are proposed and first results are presented.

Applications for a people detection and tracking algorithm using a time-of-flight camera

This paper outlines a method and applications for detection and tracking of people in depth images, acquired with a low-resolution Time-of-Flight (ToF) camera. This depth sensor is placed perpendicular to the ground in order to provide distance information from a top-view position. Usage of intrinsic and extrinsic camera parameters allows estimation of a ground plane and comparison to the measured distances of the ToF sensor in every pixel. Differences to the expected ground plane define foreground information, that is subsequently combined to associated regions. These regions of interest (ROI) are analyzed to distinguish persons from other objects by using a matched filter that is applied the height segmented depth information of each of these regions. The proposed method separates crowds into individuals and facilitates a multi-object tracking system based on Kalman filtering. Furthermore, we present several applications for the proposed method. Experiments with different crowding situations - from very low to very high density - and different heights of camera placements have proven the applicability and practicability of the system.

Realization of a recursive 3-D cone filter for video processing applications

In this paper a realization of a recursive 3-D cone filter applicable for low cost hardware like FPGAs, for instance, is presented. Based on the filter design method proposed by Bolle, multidimensional wave digital filters are used to approximate a cone-shaped frequency response. Focus is not set on filter design but on filter implementation and problems arising in that context. In detail, the impact of signal size, chosen processing scheme and shift operators on hardware implementation is illuminated and a suitable solution is proposed. For verification, an example of a filtered video sequence showing moving objects is given.

Laser phase shift distance meter for vision based driver assistance systems

Camera based systems are gaining popularity in automotive driver assistance systems. They are used to detect obstacles, road signs, road markings and other road users. Today, even low cost camera technology enables obtaining a very good image resolution and therefore cameras are able to give a good 2-dimensional image. Due to the projection of an image into the cameras sensor plane, its main disadvantage is the difficulty to estimate the distance of objects projected onto the sensor plane. This difficulty can be overcome if image information is exchanged by several cars via car-to-car-communication. Thinkable solutions range from simply exchanging estimated positions of detected objects up to tomographic image processing. For these methods, knowledge of an exact position of the vehicles is crucial. Conventional positioning methods do not deliver the required accuracy. Therefore a phase shift laser distance measuring device is proposed that is used to measure the vehicles precise position relative to landmarks. Considering the special requirements resulting from this application, the new laser measurement system design is introduced. Measurement results demonstrate its applicability. Special attention is laid on error sources such as crosstalk in this new implementation of a phase shift laser range finder.

On the usability of practical stable n-D systems for signal processing applications

In practice, n-D signal processing problems often show the property that underlying signals are unbounded with respect to only one direction, like the spatio-temporal signal of a line-scan camera, for example. Thus, the use of systems which meet the conditions of practical BIBO stability, mainly known from the field of control theory, seems to be feasible. In this paper the practical BIBO stability concept introduced by Agathoklis and Bruton [1] is analyzed with respect to its applicability in the field of signal processing applications. It is shown that practical stability in its original form is not sufficient in signal processing if not further conditions are supposed. This is done by comparing the 2D Fourier transform of the actually measured impulse response of exemplary systems to the frequency response expected on basis of the transfer function.

FPGA-implementation of signal processing algorithms for video based industrial safety applications

Conventional protective devices as light curtains allow safe but often inconvenient flow of work. Unfortunately uncomfortable safety devices are often bypassed or simply switched off. Consequently, the design of a video based protective device, which avoids inconvenient processing steps is of special interest. The present paper describes favorable combinations of FPGA-hardware and algorithms, which allow safeguarding of work places if several constraints are met. The methods were originally developed for surveillance of press brakes, but it is easily adaptable to different types of machines or work places.

Insufficiencies of practical BIBO stable n-D systems

This paper is focused on analyzing the practical bounded input bounded output (BIBO) stability concept originally introduced by Agathoklis and Bruton with respect to its applicability in n-D signal processing applications. By comparing the frequency response expected on basis of the transfer function to the Fourier transform of the actually measured impulse response of exemplary systems it is shown that the concept of practical stability cannot be a general n-D stability criterion for systems, in particular for signal processing applications, if roots of the given transfer function are not guaranteed to be close to the respective n-D BIBO stability region.

People Detection and Tracking from a Top-View Position Using a Time-of-Flight Camera

This paper outlines a method for the detection and tracking of people in depth images, acquired with a low-resolution Time-of-Flight (ToF) camera. This depth sensor is placed perpendicular to the ground in order to provide distance information from a top-view position.

On multidimensional velocity filter banks for video-based motion analysis of world-coordinate objects

Linear multidimensional filter theory provides an elegant way to analyse movement within image sequences, i.e. to separate objects moving in different directions or at different speeds. Due to its time-invariant transfer function a single filter is restricted to specific motions, for which reason a consecutively parameterized filter bank is needed to differentiate a broader set of movement. In this paper, an introduction to velocity filtering is given and unwanted side-effects are discussed, i.e. spatio-temporal ringing and perspective distortion. As a practical example a traffic scene is analysed by means of inverse perspective mapping followed by a velocity filter bank.