# US 2019/0172345 A1 · System and method for detecting dangerous vehicle > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2019/0172345 A1 - Document ID: US20190172345A1 - Status: Published application - Kind: A1 - Application number: 15/834,032 - Filed: 2017-12-06 - Published: 2019-06-06 - Earliest claimed: 2017-12-04 - Assignees: Institute for Information Industry - Applicants: Institute for Information Industry - Inventors: Chi-Sheng LIN; Chien LEE; Wei-Lun HSIAO; Yu-Ting HSU - Web page: https://patentgrep.com/patent/US20190172345A1 ## CPC - G08G 1/0175: {by photographing vehicles, e.g. when violating traffic rules} - G08G 1/0116: {from roadside infrastructure, e.g. beacons} - G08G 1/0133: {for classifying traffic situation} - G08G 1/0141: {for traffic information dissemination} - G08G 1/015: with provision for distinguishing between two or more types of vehicles, e.g. between motor-cars and cycles - G08G 1/04: using optical or ultrasonic detectors - G08G 1/091: {Traffic information broadcasting (broadcasting communication H04H)} - G08G: TRAFFIC CONTROL SYSTEMS (guiding railway traffic, ensuring the safety of railway traffic B61L; arrangement of road signs or traffic signals E01F9/00; radar or analogous systems, sonar systems, lidar systems specially adapted for traffic control G01S13/91, G01S15/88, G01S17/88; {radar or analogous systems, sonar systems, lidar systems specially adapted for anti-collision purposes G01S13/93, G01S15/93, G01S17/93}) ## Abstract The present disclosure provides a system and a method for detecting a dangerous vehicle. This method includes steps as follows. Vehicle detectors spaced apart from each other are provided, and each vehicle detector obtains a traffic image. The server infers the interaction among the vehicles in the traffic image according to a car-following theory, so as to find at least one outlier vehicle from the vehicles and to select the outlier vehicle as a focus vehicle to be tracked. The server determines whether the driving behavior of the focus vehicle falls into an abnormal behavior model. ## Claims ### Claim 1 A system for detecting dangerous vehicle, the system comprising: a plurality of vehicle detectors spaced apart from each other, and each vehicle detector configured to obtain a traffic image; and a server communicated with the vehicle detectors, and the server configured to infer an interaction among vehicles in the traffic image according to a car-following theory, so as to find at least one outlier vehicle from the vehicles and to select the outlier vehicle as a focus vehicle to be tracked, and the server configured to determine whether a driving behavior of the focus vehicle falls into an abnormal behavior model. ### Claim 2 (depends on 1) The system of claim 1, wherein the server determines a size and a moving direction of a focus area surroundings the focus vehicle required to be detected according to a direction, a speed and a position of the focus vehicle. ### Claim 3 (depends on 1) The system of claim 1, wherein the server recognizes types of the vehicles from the traffic image. ### Claim 4 (depends on 1) The system of claim 1, wherein the server collects driving track data of the vehicles from the traffic image. ### Claim 5 (depends on 1) The system of claim 1, wherein the abnormal behavior model comprises a violation condition of a plurality of traffic rules, and when the server determines that the driving behavior of the focus vehicle violates at least one of the traffic rules, the server determines that the focus vehicle is abnormal. ### Claim 6 (depends on 1) The system of claim 1, wherein the abnormal behavior model comprises at least one abnormal track, when the server determines that a driving track of the focus vehicle is different from driving tracks of others of the vehicles, and when the driving track of the focus vehicle meets the at least one abnormal track, the server determines that the focus vehicle is abnormal. ### Claim 7 (depends on 1) The system of claim 1, wherein the abnormal behavior model includes at least one abnormal speed difference range, the server compares a speed of the focus vehicle with an average speed of others of the vehicles, and when a speed difference between the driving speed of the focus vehicle and the average driving speed falls within the at least one abnormal speed difference range, the server determines that the focus vehicle is abnormal. ### Claim 8 (depends on 1) The system of claim 1, wherein the abnormal behavior model includes at least one abnormal distance, and when the server determines that a distance between the focus vehicle and any of others of the vehicles is less than the at least one abnormality distance, the server determines that the focus vehicle is abnormal. ### Claim 9 (depends on 1) The system of claim 1, wherein the server performs an alert processing procedure after the driving behavior of the focus vehicle has fallen into the abnormal behavior model. ### Claim 10 (depends on 1) The system of claim 1, wherein each of the vehicle detectors is a roadside camera. ### Claim 11 A method for detecting a dangerous vehicle, the method comprising steps of: providing a plurality of vehicle detectors spaced apart from each other, and each vehicle detector configured to obtain a traffic image; and using a server configured to infer an interaction among vehicles in the traffic image according to a car-following theory, so as to find at least one outlier vehicle from the vehicles and to select the outlier vehicle as a focus vehicle to be tracked, and the server configured to determine whether a driving behavior of the focus vehicle falls into an abnormal behavior model. ### Claim 12 (depends on 1) The method of claim 1, further comprising: using the server to determine a size and a moving direction of a focus area surroundings the focus vehicle required to be detected according to a direction, a speed and a position of the focus vehicle. ### Claim 13 (depends on 11) The method of claim 11, further comprising: using the server to recognize types of the vehicles from the traffic image. ### Claim 14 (depends on 11) The method of claim 11, further comprising: using the server to collect driving track data of the vehicles from the traffic image. ### Claim 15 (depends on 11) The method of claim 11, wherein the abnormal behavior model comprises a violation condition of a plurality of traffic rules, and the method further comprises: when the server determines that the driving behavior of the focus vehicle violates at least one of the traffic rules, determining that the focus vehicle is abnormal by using the server. ### Claim 16 (depends on 11) The method of claim 11, wherein the abnormal behavior model comprises at least one abnormal track, and the method further comprises: when the server determines that a driving track of the focus vehicle is different from driving tracks of others of the vehicles, and when the driving track of the focus vehicle meets the at least one abnormal track, determining that the focus vehicle is abnormal by using the server. ### Claim 17 (depends on 11) The method of claim 11, wherein the abnormal behavior model includes at least one abnormal speed difference range, and the method further comprises: using the server compares a speed of the focus vehicle with an average speed of others of the vehicles; and when a speed difference between the driving speed of the focus vehicle and the average driving speed falls within the at least one abnormal speed difference range, determining that the focus vehicle is abnormal by using the server. ### Claim 18 (depends on 11) The method of claim 11, wherein the abnormal behavior model includes at least one abnormal distance, and the method further comprises: when the server determines that a distance between the focus vehicle and any of others of the vehicles is less than the at least one abnormality distance, the server determines that the focus vehicle is abnormal. ### Claim 19 (depends on 11) The method of claim 11, further comprising: using the server performs an alert processing procedure after the driving behavior of the focus vehicle has fallen into the abnormal behavior model. ### Claim 20 (depends on 11) The method of claim 11, wherein each of the vehicle detectors is a roadside camera.