Image processing is widely considered an essential part of future driver assistance systems. This paper presents a motionbased vision approach to initial detection of static and moving objects observed by a monocular camera attached ,to a moving observer. The underlylng principle is based on parallax flow induced by all non-planar static or moving object of a 3D scene that is determined from optical Row measurements. Initial object hypotheses are crested in regions containing significant parallax flow. The significance is determined from planar parallax decomposition automatically. Furthermore, we propose a separation of detected image motion into three hypotheses classes, namely coplanar, static and moving regions. To achive a high degree of robustness and accuracy in real traffic situations some key processing steps are supported by the data of inertial sensors rigidly attached to our vehicle. The proposed method serves as a visual short-range surveillance module providing instantaneous object candidates to a driver assistance system. Our experiments and simutations confirm the feasibility and robustness of the detection method even in complex urban environment. I, INTRODUCTIONVisual detection is one of the most challenging issues in computer vision. Numerous solutions and system realizations have been published which cover a wide range of applications including surveillance and security systems, vehicle detection as well as applications in manufacturing and robotics. Our paper describes an image processing method suited for visual surveillance supporting new functions in driver assistance systems such as, enhanced radar-based adaptive cruise control with full speed range (ACC-FSR), stop&go or lane keeping support. In a typical sensor configuration the field of view of the video sensor will be significantly broader than those or radar sensors. While cooperative sensing techniques can be employed in the region under surveillance of multiple sensors, one is left with lateral regions that are only viewed by the video sensor.Our method visually detects and segments static and moving objects, i.e. potential obstacles and other moving traffic members. It is intended to work in vicinity to the ego-vehicle.AS a concise objective of this contribution we focus on the detection of cut-in and overtaking vehicles. The limitation to short-range surveillance is mainly due to employment of the motion feature for detection. The differences in image motion between foreground and background decrease for increasing distance. This fact limits the range of reliably detectable motion information under real conditions. Many different solutions have been proposed to tackle the problem of initial detection of scene objects while the observing camera itself is moving. In such cases difference-based methods fail to detect moving objects due to the apparent motion of the scene induced by the observer's motion, which we call ego-motion in the sequel. In this paper we propose a detection scheme where a set of measured visual motion features ...

The paper reports on investigations concerning the application of block oriented fractal coding schemes for encoding of color images. Correlations between the different color planes can be exploited for the aim of data compression. For this purpose the similarities between the fractal transform parameters of one block location but different color planes aie regarded in a blockwise manner. Starting-point is the fractal code for one block in the dominant color plane which serves as prediction for the code of the consponding block in the other planes. Emeiging from this prediction the depending codes can be derived by a successive refinement strategy. Since the fractal code for the dominant color plane and the refinement information for determining the code for the other planes can be represented with fewer bits compared to the independently calculated ones, a coding gain can be achieved.

No abstract

No abstract

No abstract