Ritesh Kanjee

Abstract—Adaptive Cruise Control (ACC) is a relatively new system
designed to assist automobile drivers in maintaining a safe following
distance. This paper proposes and validates a vision-based ACC system
which uses a single camera to obtain the clearance distance between the
preceding vehicle and the ACC vehicle. Pattern matching, with the aid
of lane detection, is used for vehicle detection. The vehicle and range
detection algorithms are validated using real-world data, and… Show more

Abstract—Adaptive Cruise Control (ACC) is a relatively new system
designed to assist automobile drivers in maintaining a safe following
distance. This paper proposes and validates a vision-based ACC system
which uses a single camera to obtain the clearance distance between the
preceding vehicle and the ACC vehicle. Pattern matching, with the aid
of lane detection, is used for vehicle detection. The vehicle and range
detection algorithms are validated using real-world data, and then the
resulting system performance is shown to be sufficient using a simulation
of a basic vehicle model.

Index Terms—Vision-Based Adaptive Cruise Control, Adaptive Image
Cropping, PI Controller, Autonomous Vehicle, Fiducial Detection, Pattern
Matching, Single Camera.
Show less

This paper proposes and validates a real-time on-road vehicle detection system, which uses a single camera for the purpose of intelligent driver assistance. A three-step vehicle detection framework is presented to detect and track the target vehicle within an image. In the first step, probable vehicle locations are hypothesized using pattern recognition. The vehicle candidates are then verified in the hypothesis verification step. In this step, lane detection is used to filter vehicle… Show more

This paper proposes and validates a real-time on-road vehicle detection system, which uses a single camera for the purpose of intelligent driver assistance. A three-step vehicle detection framework is presented to detect and track the target vehicle within an image. In the first step, probable vehicle locations are hypothesized using pattern recognition. The vehicle candidates are then verified in the hypothesis verification step. In this step, lane detection is used to filter vehicle candidates that are not within the lane region of interest. In the final step tracking and online learning are implemented to optimize the detection algorithm during misdetection and temporary occlusion. Good detection performance and accuracy was observed in highway driving environments with minimal shadows. Show less