A map-matching method based on the filtering of parking point data of floating vehicles comprises the steps of: step one, establishing a floating vehicle information database; step two, separating thefloating vehicle data record at the moving points from the floating vehicle data record at the parking points; step three, calculating the projection distance of the moving points of the floating vehicles at the moving points and the included angle in the driving direction; step four, calculating the distance measurement value of each floating vehicle at the moving point; step five, filtering theroad section where the floating vehicles at the moving points are located; step six, calculating the projection distance of the parking points of the floating vehicles at the parking points; step seven, filtering the road section where the floating vehicles at the parking points is located; step eight, initially judging the road section where the floating vehicles at the parking points is located; step nine, setting the length of a queue for the floating vehicles at the parking points; step ten, calculating the distance from the floating vehicles at the parking points to the terminal of the road section; step eleven, judging the road section where the floating vehicles at the parking points is located for a second time; and step twelve, judging the road section where the floating vehiclesat the parking points is located for a third time. The method can match the zero velocity point which can show the traffic state of the road section with the road section.