System and method for measuring the real traffic flow of an area

• US 9,390,621 B2
• Filed: 06/03/2015
• Issued: 07/12/2016
• Est. Priority Date: 06/04/2014
• Status: Active Grant

• Current Assignee: Cuende Infometrics SA
• Original Assignee: Cuende Infometrics SA
• Inventors: Cuende Alonso, Daniel
• Primary Examiner(s): Patel, Nirav G
• Assistant Examiner(s): SHIN, BRIAN D
• Application Number:
  US14/729,092
• Publication Number:
  US 20150356868A1
• Time in Patent Office: 405 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G01C 21/36 Input/output arrangements for on-board computers
• CPC Class Codes: G06V 20/176 Urban or other man-made structures
• CPC Class Codes: G08G 1/0104 Measuring and analyzing of parameters relative to traffic conditions
• CPC Class Codes: G08G 1/0125 Traffic data processing
• CPC Class Codes: G08G 1/04 using optical or ultrasonic detectors

Abstract

A system and method for measuring a real traffic flow of an area use images of vehicles and pedestrians taken from the sky in time intervals of an area. The system and method also use continuous tracking data for vehicles and pedestrians, provided by navigator manufacturers, map providers, route applications or by phone carriers. Both images and continuous tracking data are combined and processed by a processor that applies image pattern recognition software to the images so that each vehicle and pedestrian in the images is recognized and labeled for which tracking data is available. The system identifies vehicles and pedestrians contained in both databases and applies a weight to thereto. Accordingly, the system and method are able to estimate the traffic flow and the total number of vehicles and pedestrians for a particular time and area.

Claims

1. A system for measuring a real traffic flow of an area, the system comprising:

images taken from a sky in time intervals of an area;

continuous tracking data for a probe that is collected and provided by navigator manufacturers, map providers, route applications or by phone carriers that record cell towers, traffic counts, router connections, near-field-communication devices, and beacons that are able to geolocate the probe;

the continuous tracking data comprising at least a longitude, latitude for each time unit for which the probe is tracked;

a database of universe of elements that comprises the images;

a continuous database that comprises the continuous tracking data; and

a computer processor configured to;

apply image pattern recognition software to the images taken from the sky so that each element comprised in the images is recognized, labeled and at least a latitude, longitude, time, speed and type of vehicle or number of pedestrians is calculated;

allocate to each probe the continuous tracking data for which the longitude, latitude and time match;

calculate a ratio between a number of elements recognized in the images and a number of probes for which the longitude, latitude and time match; and

estimate the flow and the total number of elements within an area by applying the calculated ratio to the continuous tracking data for a particular time for which an image from the sky is not available.

2. The system of claim 1, wherein the computer processor is further configured to:

calculate a pattern for each probe based on the continuous tracking data; and

calculate a profile by analyzing combinations of the calculated patterns.

3. The system of claim 1, wherein the images are provided by at least one aerial device.

4. The system of claim 1, wherein the at least one aerial device is selected from a satellite, a plane, an aircraft, a drone and any combination thereof.

5. The system of claim 1, wherein the probe is selected from a tracked vehicle and a tracked pedestrian.

6. The system of claim 1, wherein the area is selected from a street, a suburb, a city, a town, a country and a continent.

7. The system of claim 1, wherein the images include at least one of photos and moving frames.

8. A method for measuring a real traffic flow of an area, the method comprising:

generating a database of universes which stores images taken from a sky in time intervals of an area;

recognizing, using a processor, each individual element comprised in each stored image and calculating at least the following data of each individual element;

latitude, longitude, time, speed and type of vehicle or number of pedestrians;

generating a continuous database which stores continuous tracking data of each individual probe, the continuous tracking data comprising at least an estimated position by longitude and latitude, map cell and vector position, for each time unit;

the continuous tracking data being obtained from each tracking device carried by each individual probe;

allocating, to each individual element recognized in the step of recognizing each individual element comprised in each stored image, the continuous tracking data for which the longitude, latitude and time match;

calculating a ratio between the number of elements recognized in the images and the number of probes for which the longitude, latitude and time match; and

estimating the flow and the total number of elements within an area by applying the calculated ratio to the continuous tracking data for a particular time for which an image from the sky is not available.

9. The method of claim 8, wherein the method further comprises:

calculating a pattern for each probe based on the continuous tracking data; and

calculating a profile by analyzing combinations of the calculated patterns.

10. The method of claim 8, wherein generating the database of universes further comprises:

defining a Universe region which is a composition of several images;

selecting different source data of aerial images selected from;

a satellite, airplanes and drones, with latitude and longitude and a synchronized timestamp;

normalizing the input image;

generating a chronogram of the different images and regions to sync later with a snapshot of the continuous database;

counting vehicle and pedestrian analysis;

correcting and enhancing the data; and

,

counting elements.

11. The method of claim 8, wherein the step of counting vehicle and pedestrian analysis further comprises:

mapping correlating of roads wherein roads and car parks are placed as vectors in the images to highlight where it is possible to find vehicles;

calibrating, wherein objects are defined for the image recognition, the objects being at least;

car, type, model, bus, truck, motorbike, bicycle, bus shelter, billboard, kiosk, urban furniture and pedestrian; and

processing the images with image pattern recognition software based on a visible range, infrared and ultraviolet.

12. The method of claim 8, wherein the step of correcting and enhancing the data further comprises:

matching to vector navigation map;

correcting an effect of traffic lights or traffic jams for possible false positives as parked cars;

deducing a movement and trajectory for each individual element before a snapshot and after the snapshot;

forecasting the movement for each individual element based on type of road, type of element and speed; and

identifying regions and locations to be excluded or requiring a different analysis.

13. The method of claim 8, wherein the step of counting elements further comprises:

counting by segment of a street, number of vehicles, cars, buses, motorcycles, moving and parked, people and static probes for each image and for total images in a period of time; and

counting a total for the area for each image and for total images in a period of time.

14. A method for measuring a real traffic flow of an area, the method comprising:

generating a database of universes which stores images taken from a sky;

normalizing the stored images of the database of universes;

looking, using a computer processor, for elements of a continuous database which were in movement at a time of one of the stored images of the database of universes;

assigning, using the computer processor, a weight to each probe of the continuous database in terms of elements with matching; and

repeating, using the computer processor, the steps of looking for elements of the continuous database which were in movement at the time of one of the stored images of the database of universes and assigning the weight to each probe of the continuous database in terms of elements with matching, for each stored image, and combining the weights obtained;

wherein the combination of weights takes into account at least whether it is a weekend or a working day, summer or winter; and

wherein the weights are distributed among continuous database probes that never match with the database of universes elements by matching with other continuous database probes already weighted.