A Real-time 2-Dimentional Recovery Algorithm for Blurred Video Image Filmed by Charge-Coupled Devices in Environment Sensing for Driving Safety

Enhancing driving safety is always a thorny issue for transportation management department in any cities, for road traffic environment in most cases can be complex and full of uncertainties. As traffic accidents are always caused by moving disturbing objects (including people) which may not be perceived by vehicle drivers, those disturbing objects can usually be filmed as blurred video images by certain equipments such as driving recorders. Only if those images be processed on time properly by image processing techniques, can the real-time driving environment be sensed and can the drivers be helped. Till now, most related works have been reported to solve blurred video image recovery problem via processing it offline, and the only online approaches are somehow complex or not comprehensive in computational process for moving objects, and are not proposed for the very purpose of enhancing driving safety. In this paper, a real-time, 2-dimensional recovery algorithm for blurred video image filmed by charge-coupled devices (CCD) in driving environment sensing with high computational efficiency has been proposed, the algorithm has been described in detail via mathematical derivation, and a case study has been done to verify the effectiveness of it.