Moheb Mekhaiel

A pedestrian detection system based on thermopile and radar sensor data fusion

Automotive pedestrian protection systems will be introduced in the EU in short term to reduce the number of accidents and injury fatalities. As with any safety issue, a comprehensive approach comprising both active and passive safety elements should be followed and it is also valid for pedestrian protection. Passive safety short term solutions can be contact sensor systems that trigger raisable engine hoods. This paper discusses an innovative approach for pedestrian detection and localization, by presenting a system based on two short range radars and an array of passive infrared thermopile sensors, aided with probabilistic techniques for detection improvement. The two short range radars are integrated in the front bumper of the test vehicle. They are able to observe and track multiple targets in the region of interest. However, one difficulty is to distinguish between pedestrians and other objects. Therefore, a second sensor system is required to classify pedestrians reliably. This system consists of spatial distributed thermopile sensors which measure the object presence within their respective field-of-view independently. These measurements are then validated and fused using a mathematical framework. Thermopiles are excellent to detect the thermal radiation emitted by every human. However, a robust signal-interpretation algorithm is mandatory. In this work a statistical approach combining Dempster-Shafer theory with occupancy-grid method is used to achieve reliable pedestrian detection. Thermopile and radar sensors use independent signature-generation phenomena to develop information about the identity of objects within the field of view. They derive object signatures from different physical processes and generally do not cause a false alarm on the same artifacts. The integration of the sensor readings from the radar and thermopile system is conducted using unifying sensor-level fusion architecture.

Grid-based optimal sensor arrangement within a sensor array for 2D position estimation

In this paper, an innovative grid-based approach to optimize the sensor arrangement in a sensor array is presented. Such an optimized sensor arrangement involving low-cost sensors finds its use in a variety of areas as in automotive safety applications. The potential cost-savings achievable through the use of such sensor systems makes them attractive for car makers. The presented approach is based on an array of N sensors located in a horizontal plane. The sensors that are being used are low-cost infrared sensors, which provide an output voltage depending on the presence of an object inside the sensors field of view (FOV). Since every sensor has a limited FOV and range-of-sight, several sensors with overlapping FOVs are necessary to cover any specific region of interest (ROI) satisfactorily. The goal of the presented approach is to identify an optimal sensor arrangement for object localization within the ROI. Since only the signal changes are used for processing and not their absolute values, it is imperative that FOVs of at least two sensors should overlap for object position estimation. In addition, the accuracy of the estimation depends on the size of the overlapping area. To solve this (multi-faceted) optimization problem there were no convenient analytic solutions available. Therefore a numerical grid-based solution to compute the best sensor arrangement was developed in this approach. The ROI is represented by a regular grid. Every cell in the grid has a determined weighting function. The formulation of the weighting function is the key to the optimization problem and it is based on the following parameters: number of sensors that cover the cell, dimensions of the overlapping areas and the complete coverage of the ROI. The sum over all cell-weights within the grid is the cost function to be optimized. Due to the high computational effort several algorithms are considered and the final implementation with a simulated annealing approach is chosen because of its ability to find a global minimum in reasonable time.

Fusion von Radar- und Thermopilesensordaten zur Fußgängerdetektion (Fusion of Radar and Thermopile Sensor Data for Pedestrian Detection)

Viele fahrzeugtechnische Innovationen der letzten 30 Jahre zielten auf die Vermeidung oder Verringerung von Personenschäden bei Unfällen. Die sogenannte passive Sicherheit hat zu einem großen Teil dazu beigetragen, dass die Zahl tödlicher Verletzungen der Fahrzeuginsassen stark zurückgegangen ist. Mehr noch als bisher ist die Automobilindustrie daran interessiert, die aktiven Sicherheitssysteme voranzutreiben. Innovative Sicherheitssysteme, wie z.B. Fußgängerschutzsysteme, sollen in der Zukunft durch aktive Maßnahmen dafür sorgen, dass Fußgänger im Straßenverkehr besonders geschützt werden.In diesem Artikel wird ein System zur Fußgängerdetektion, basierend auf Radar- und Thermopilesensoren, vorgestellt. Hierbei liegt der Fokus vor allem auf der Sensordatenfusion mit Hilfe der Dempster–Shafer-Theorie. Mit Hilfe eines solchen Systems, das im DaimlerChrysler Forschungszentrum in Ulm entwickelt wird, soll es in der Zukunft möglich sein, die Sicherheit von Fußgängern im Straßenverkehr durch aktive Sicherheitsmaßnahmen zu erhöhen. Whenever addressing pedestrian-related injuries in automotive accidentology, a comprehensive approach comprising both active and passive safety elements should be followed. Passive safety solutions can be contact sensor systems that trigger raisable engine hoods, and an active safety element could be the Brake Assistant. However, an important enabler for a future pedestrian protection system is a suitable, low-cost, environment-friendly sensing technology for pedestrian detection, supported by a fast and reliable algorithm for object localization. This paper discusses such an innovative approach for pedestrian detection and localization, presenting a system based on both short-range radar sensors and an array of passive infrared thermopile sensors. The system has been developed at the DaimlerChrysler research center in Ulm.

Short Range Radar System for Automotive Applications

The number of cars in Europe is increasing, leading to a higher traffic density. The average age of drivers is increasing consistent with demographics of the total European population. Every second accident involving vehicles is related to traffic situations in which faster reaction of the driver could have mitigate crash consequences. So, there is an increased need and appreciation for obstacle detection that operate at day and night.