Thomas Schamm

Obstacle detection with a Photonic Mixing Device-camera in autonomous vehicles

In autonomous vehicles as well as in modern driver assistance systems, obstacle detection shows to be the most important task to be achieved. This paper presents a collision avoidance system, based on a modern Time-Of-Flight camera. These cameras allow a 3D perception of the environment, in which obstacles can be detected, independent of special features. Thus, the system is capable of all kinds of objects, including pedestrians as well as bicycles or vehicles. The used Photonic Mixing Device (PMD) camera has a measurement range of up to 50 m. The system is integrated into an autonomous vehicle, on which detected obstacles are investigated in detail. The vehicle steering commands are then generated by a behaviour network, depending on the presence of obstacles in the driving lane.

Contextual task-aware shared autonomy for assistive mobile robot teleoperation

For robot applications in unknown or even hazardous environments, such as search and rescue, it is difficult and stressful for human beings to merely simply teleoperate a mobile robot without its assistance. Consequently, means to facilitate an efficient shared autonomy between human and robot are the subject of much research work in the field of robotics. This paper proposes a novel shared autonomy system, which recognizes the user intention by estimating the task the user is executing based on the context information, and provides motion assistance according to the inferences. To incorporate the uncertainty of contextual task recognition, a Gaussian Mixture Regression model combined with a recursive Bayesian filter is adopted, which is adaptive to the implicit user model for task execution during operation. The proposed method is applied to the problem of controlling a flying robot in the context of two task types: doorway crossing and object inspection. Its benefits are demonstrated by the simulation results.

A model-based approach to probabilistic situation assessment for driver assistance systems

Until today, driver assistance systems deduce actions from very limited information, which consider the driving situation. Only few aspects of complex interactions between the driver, the vehicle and the environment are regarded, a holistic driving situation is not assessed. In this work, we present a feasible approach to model driving situations using a knowledge base. The knowledge is described by first-order logic in a formal language. Propelled by environmental information, a probabilistic network is automatically constructed from the formal definitions, to overcome the rigid nature of traditional networks. The network is continuously updated by sensor information and probabilistic inference of the situation is performed. The method proposed is eligible for driving situation assessment, which is demonstrated in examples.

Towards driving autonomously: Autonomous cruise control in urban environments

For automatic driving, vehicles must be able to recognize their environment and take control of the vehicle. The vehicle must perceive relevant objects, which includes other traffic participants as well as infrastructure information, assess the situation and generate appropriate actions. This work is a first step of integrating previous works on environment perception and situation analysis toward automatic driving strategies. We present a method for automatic cruise control of vehicles in urban environments. The longitudinal velocity is influenced by the speed limit, the curvature of the lane, the state of the next traffic light and the most relevant target on the current lane. The necessary acceleration is computed in respect to the information which is estimated by an instrumented vehicle.

Real time pedestrian detection by fusing PMD and CMOS cameras

In this work we present the preliminary results of the fusion of photonic mixer device - PMD and CMOS cameras for driver assistance applications. Although the algorithms are demonstrated mainly for pedestrians, they apply to the other objects on the street. PMD camera delivers the 3D object list. Object coordinates are further projected into CMOS image plane where classification is performed using support vector machines. As compared to PMD camera the CMOS camera has higher resolution, which gives the possibility to realize finer object detection, separation and classification. As the feature vector we use quadruple haar discrete wavelet transformation (QH DWT). The speed improvement of the SVM in the testing phase (necessary for real-time implementation) is realized with Burgpsilas reduced set vector method (BRSVM), improving classification speed nearly 70 times. We have achieved the pedestrian detection rate of 80%.

Semantic State Space for High-level Maneuver Planning in Structured Traffic Scenes

Originating from simple cruise control systems that monitor and control the speed of the vehicle, driver assistance systems have evolved into intelligent systems. Future assistance systems will combine information from different sensors and data sources to build up a model of the current traffic scene. This way they will be able to assist in challenging tasks in complex situations. Towards this goal, we present a semantic scene representation for modeling traffic scenes. Based on a geometric representation a semantic representation is defined using an ontology to model relevant traffic elements and relations. Considering potential relations of the ego vehicle, a semantic state space of the ego vehicle is derived. Transitions are defined that model state changes (maneuvers). The model can be used for example for situation analysis and high level planning for driving hint generation or automated driving. The method is evaluated in different traffic situations and on real sensor data. It is going to be applied to (semi-)automated driving in a real test vehicle.

Semivirtual simulations for the evaluation of vision-based ADAS

The design and development process of advanced driver assistance systems (ADAS) is divided into different phases, where the algorithms are implemented as a model, then as software and finally as hardware. Since it is unfeasable to simulate all possible driving situations for environmental perception and interpretation algorithms, there is still a need for expensive and time-consuming real test drives of thousands of kilometers. Therefore we present a novel approach for testing and evaluation of vision-based ADAS, where reliable simulations are fused with recorded data from test drives to provide a task-specific reference model. This approach provides ground truth with much higher reliability and reproducability than real test drives and authenticity than using pure simulations and can be applied already in early steps of the design process. We illustrate the effectiveness of our approach by testing a vision-based collision mitigation system on recordings of a german highway.

COLLISION AVOIDANCE FOR COGNITIVE AUTOMOBILES USING A 3D PMD CAMERA

Abstract Collision avoidance is one of the most important capabilities for autonomous vehicles. During driving, collisions must be avoided in all situations. With the availability of 3d cameras which rely on the time-of-flight principle, it is possible to get a very rich perception of the environment. This paper shows, how obstacles can be detected in the vehicles surrounding using a 3d PMD-camera (photonic mixing device). The obstacle detection is composed of two separated steps. First, a segmentation and a clustering of pixels takes place. Secondly, each group of pixels is analyzed in order to decide whether it is an obstacle or not. The result of the detection is a list of obstacles which is then used for behavior execution. The execution is done with a behavior network and it generates recommendations for path planning.

Anticipatory energy saving assistant for approaching slower vehicles

Due to scarcity of petrol and emission reduction there is a need for more energy efficiency in individual transportation. Beside optimizations of carriage, engine and power train, anticipatory driving plays a key roll for efficient driving. Advanced driver assistance systems can help the driver in this task. However, current assistance systems do not provide timed hints to the driver. In this work the development of a system is presented that assists the driver to save energy while approaching slower vehicles. Data from several sources like static maps, vehicle data and sensor data is used to generate hints. These are displayed to the driver in a graphical user interface. System performance is evaluated on a simulated test-run.

Testing and validating high level components for automated driving: simulation framework for traffic scenarios

Current advances in the research field of autonomous driving demand advanced simulation methods for testing and validation. By combining versatile foci of different simulations, we can provide an increased amount and diversity of realistic traffic scenarios, which are relevant to the development and verification of high level automated driving functions. The focus of the present paper is to propose a concept for realistic simulation scenarios, which is capable of running in different integration levels, from software- to vehicle-in-the-loop. Its application is demonstrated, exposing an experimental vehicle, which is used for autonomous driving development, to a traffic scenario with virtual vehicles on a real road network.