Methods and systems for moving traffic obstacle detection

Patent metadata

Field	Value
Publication number	US-2018121741-A1
Application number	US-201715859382-A
Country	US
Kind code	A1
Filing date	Dec 30, 2017
Priority date	Jun 9, 2016
Publication date	May 3, 2018
Grant date	—

Abstract

The disclosure provides systems and methods for detecting, characterizing, and predicting moving traffic obstacles. The systems and methods are suitable for densely populated areas in resource-constrained regions. With the characterizations and predictions of moving traffic obstacles, a variety of benefits can accrue to individuals and devices that use traffic information.

First claim

What is claimed is:

1 . A method for maneuvering a vehicle in a roadway having potential traffic obstacles, the method comprising:
establishing a communications link between a server and a device in a vehicle; and communicating an alert to the device, wherein the alert is configured to cause the vehicle to maneuver based on a potential moving traffic obstacle path, wherein the potential moving traffic obstacle path is determined based on sensor data for a road segment, the sensor data corresponding to observations of an obstacle at a location along the road segment over a period of time.

2 . The method of claim 1 , wherein the sensor data is obtained from one or more sensors selected from: an onboard sensor on one or more probe vehicles as the one or more probe vehicles traverses the road segment; a fixed location sensor monitoring the road segment; and a combination thereof.

3 . The method of claim 1 , wherein the sensor data corresponds to observations of one or more obstacles at one or more locations along the road segment over a period of time.

4 . The method of claim 1 , wherein the alert is generated when the vehicle comes into proximity of the potential moving traffic obstacle path, and wherein the alert is configured to alter a user interface to display information suitable to cause a driver to initiate a maneuver, or wherein the alert is configured to cause an autonomous vehicle control system to initiate a maneuver.

5 . The method of claim 1 , wherein determination of the potential moving traffic obstacle path involves:
receiving the sensor data for the road segment by a server via a network; determining a set of gray scale pixels potentially missing based on the observations, each pixel representing a time and location combination, and having an associated obstacle probability assigned based on the sensor data; applying a modified Probabilistic Hough Transform to the set of gray scale pixels to obtain a Hough Transform space, wherein each value in the Hough Transform space is a log likelihood supported or opposed by each observation; and identifying the potential moving traffic obstacle path by identifying a linear arrangement of locally maximum peak values in the Hough Transform space.

6 . The method of claim 1 , wherein the potential moving traffic obstacle path is stored in a potential moving traffic obstacle path database.

7 . The method of claim 1 , wherein the communication is configured to cause a change in a graphical user interface on the device such that a user is notified of the potential moving traffic obstacle and is further notified of a suggested maneuver based on the potential moving traffic obstacle.

8 . The method of claim 1 , wherein the communication is configured to cause an autonomous vehicle control system to maneuver based on the potential moving traffic obstacle.

9 . The method of claim 1 , wherein the sensor data originates from one or more vehicles traversing the road segment, and wherein the one or more vehicles is a dedicated probe vehicle or is a non-dedicated vehicle having a mobile device configured to collect and transmit the sensor data.

10 . The method of claim 1 , wherein the sensor data is from a camera, accelerometer, gyroscope, or combination thereof.

11 . A method for maneuvering a vehicle in a roadway having potential traffic obstacles, the method comprising:
receiving sensor data for a road segment, the sensor data corresponding to observations of an obstacle at a location along the road segment over a period of time; determining a set of gray scale pixels potentially missing based on the observations, each pixel representing a time and location combination, and having an associated obstacle probability assigned based on the sensor data; applying a modified Probabilistic Hough Transform to the set of gray scale pixels to obtain a Hough Transform space, wherein each value in the Hough Transform space is a log likelihood supported or opposed by each observation; identifying a potential moving traffic obstacle path by identifying a linear arrangement of locally maximum peak values in the Hough Transform space; adding the potential moving traffic obstacle path to a potential moving traffic obstacle database; communicating an alert to a device in a vehicle, wherein the communication comprises the potential moving traffic obstacle path and is configured to cause the vehicle to maneuver based on the potential moving traffic obstacle path.

12 . A method for instructing a vehicle to maneuver in a roadway, the method comprising:
processing, by a server, sensor data received by the server from at least one device comprising a sensor, the sensor data corresponding to observations of an obstacle at a location along a road segment over a period of time, the processing comprising the steps: (a) determining a set of gray scale pixels potentially missing based on the observations, each pixel representing a time and location combination, and having an associated obstacle probability assigned based on the sensor data; (b) applying a modified Probabilistic Hough Transform to the set of gray scale pixels to obtain a Hough Transform space, wherein each value in the Hough Transform space is a log likelihood supported or opposed by each observation; and (c) identifying a potential moving traffic obstacle path by identifying a linear arrangement of locally maximum peak values in the Hough Transform space; establishing a communications link between the server and a device onboard a vehicle, the vehicle located on the road segment; transmitting to the device via the communications link a message comprising information about the potential moving traffic obstacle and instructions for avoiding the potential moving traffic obstacle, wherein the message is configured to alter a graphical user interface of the device or to automatically alter the course of the vehicle.

13 . The method of claim 12 , wherein the instructions comprise route information configured for display on a GUI and configured to instruct a user to maneuver the vehicle based on the potential moving traffic obstacle.

14 . The method of claim 12 , wherein the instructions comprise machine-readable instructions configured to cause an autonomous vehicle control system to maneuver the vehicle based on the potential moving traffic obstacle.

15 . The method of claim 12 , wherein the device comprising a sensor is a fixed device monitoring the road segment or is a mobile device configured to be disposed within a vehicle traversing the road segment.

16 . The method of claim 12 , further comprising identifying a plurality of potential moving traffic obstacle paths by identifying a plurality of linear arrangements of locally maximum peak values in the Hough Transform space and consolidating the plurality of potential moving traffic obstacle paths.

Assignees

• IBM

Inventors

• Tatsubori Michiaki

Classifications

• G08G1/0112 Primary

  from the vehicle, e.g. floating car data [FCD] · CPC title

• G06V20/58

  Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads · CPC title

• G06V10/48

  by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation · CPC title

• G05D1/0238

  using obstacle or wall sensors (G05D1/0246 and G05D1/0289 take precedence; lidar systems designed for anti-collision purposes G01S17/93) · CPC title

• G05D1/0088

  characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours (using knowledge based models G06N5/00) · CPC title