Driving safety system and barrier screening method thereof

Patent metadata

Field	Value
Publication number	US-9478139-B2
Application number	US-201414583134-A
Country	US
Kind code	B2
Filing date	Dec 25, 2014
Priority date	Dec 25, 2014
Publication date	Oct 25, 2016
Grant date	Oct 25, 2016

Abstract

A driving safety system mounted in a vehicle includes an image acquisition module and a distance detection module connected to a vehicle-mounted computer and respectively acquiring at least one piece of barrier information and at least one frame of image information. The vehicle-mounted computer analyzes the barrier information and the image information, constantly receives barrier information in the course of driving, and performs a screen and analysis algorithm to screen out noise from the surrounding environment and a comparison analysis algorithm to eliminate noise reflected from objects on the ground, to accurately acquire front target information. Accordingly, the driving safety system can correctly determine barriers in the front when the vehicle is moving to achieve the goal of enhancing driving safety.

First claim

What is claimed is:

1. A barrier screening method comprising:
mounting a driving safety system in a vehicle, with the driving safety system including a computer, an image acquisition module and a distance detection module, with the computer connected to the image acquisition module and the distance detection module; and
the computer performing the method comprising:
receiving at least one piece of barrier information of at least one barrier from the distance detection module and at least one frame of image information from the image acquisition module with each frame of image information containing the at least one barrier;
performing a screen and analysis algorithm to process each one of the at least one piece of barrier information to screen out noise existing in the at least one piece of barrier information coming from any non-barrier and reflected from a surrounding environment;
generating location information of a corresponding barrier after screening out the noise from the at least one piece of barrier information;
comparing the location information of the corresponding barrier with the at least one frame of image information to eliminate the noise generated from objects on the ground and to obtain front target information;
detecting a current driving status of the vehicle by a vehicle status module; and
sending a brake activation signal to slow down the vehicle or a brake fluid cutoff signal to a brake control module according to a determined analysis based upon the at least one piece of barrier information of the at least one barrier.

2. The barrier screening method of claim 1 , wherein in performing the screen and analysis algorithm, the computer performs a first noise-screening method, with the first noise-screening method comprising:
calculating distance information and estimated distance information of the at least one barrier;
determining when a variance fluctuates with variation in a variance value outside a fluctuation range; and
determining that the at least one piece of barrier information of the at least one barrier is from the noise when the variance fluctuates with variation in the variance value outside the fluctuation range.

3. The barrier screening method of claim 2 , wherein in performing a screen and analysis algorithm, the computer further performs a second noise-screening method, wherein the at least one barrier comprises a first barrier and a second barrier, wherein the second noise-screening method comprises:
acquiring location information of the first barrier and the second barrier;
determining if a ratio of a distance from the location information of the second barrier to the vehicle to a distance from the location information of the first barrier to the vehicle meets a preset value;
determining when the distance of the first barrier and the distance of the second barrier fall within a distance range when the ratio meets the preset value; and
determining that the location information of the second barrier is from the noise caused by a second harmonic generation from the second barrier.

4. The barrier screening method of claim 3 , wherein in comparing the location information of the corresponding barrier, the computer performs a ground noise screening method according to a first distance between the location information of the first barrier and the vehicle and a second distance between the location information of the second barrier and the vehicle, wherein the ground noise screening method comprises:
locking the first barrier as a target;
keeping track of the location information of the first barrier to generate estimated location information;
comparing the estimated location information with the location information of the first barrier subsequently acquired to obtain an absolute value;
determining when the absolute value is, less than an error threshold; and
determining that the location information of the first barrier is from the noise when the absolute value is less than an error threshold.

5. The barrier screening method of claim 4 , with the computer acquiring driving status information and a piece of barrier information and calculating a barrier speed using the driving status information and the piece of barrier information, and when the barrier speed exceeds a speed threshold, the at least one barrier is determined as a dynamic barrier, otherwise, the at least one barrier is determined as a static barrier.

6. The barrier screening method of claim 4 , with the computer identifying one of the at least one piece of barrier information most adjacent to a vehicle moving path through the at least one piece of barrier information and the at least one frame of image information and determining when the identified piece of barrier information is from a static barrier, and when the identified piece of barrier information is from the static barrier, the computer further performing vehicle recognition according to the location information of the identified piece of barrier information to ascertain if the front target information is associated with a vehicle.

7. The barrier screening method of claim 5 , with the computer identifying one of the at least one piece of barrier information most adjacent to a vehicle moving path through the at least one piece of barrier information and the at least one frame of image information and determining when the identified piece of barrier information is from a static barrier, and when the identified piece of barrier information is from the static barrier, the computer further performing vehicle recognition according to the location information of the identified piece of barrier information to ascertain if the front target information is associated with a vehicle.

8. The barrier screening method of claim 1 , with the computer acquiring driving status information and a piece of barrier information and calculating a barrier speed using the driving status information and the piece of barrier information, and when the barrier speed exceeds a speed threshold, the at least one barrier is determined as a dynamic barrier, otherwise, the at least one barrier is determined as a static barrier.

9. The barrier screening method of claim 8 , with the computer identifying one of the at least one piece of barrier information most adjacent to a vehicle moving path through the at least one piece of barrier information and the at least one frame of image information and determining when the identified piece of barrier information is from a static barrier, and when the identified piece of barrier information is from the static barrier, the computer further performing vehicle recognition according to the location information of the identified piece of barrier information to ascertain if the front target information is associated with a vehicle.

10. The barrier screening method of claim 1 , with the computer identifying one of the at least one piece of barrier information most adjacent to a vehicle moving path through the at least one piece of barrier information and the at least one frame of image information and determining when the identified piece of barrier information is from a static barrier, and when the identified piece of barrier information is from the static barrier, the computer further performing vehicle recognition according to the location information of the identified piece of barrier information to ascertain if the front target information is associated with a vehicle.

11. A driving safety system mounted in a vehicle, comprising:
a distance detection module configured to acquire at least one piece of barrier information from at least one barrier;
an image acquisition module configured to acquire at least one frame of image in

Assignees

• Automotive Res & Testing Ct

Inventors

• Hsu Ping-Min
• Li Ming-Hung

Classifications

• G06V10/811

  the classifiers operating on different input data, e.g. multi-modal recognition · CPC title

• B60T7/22 Primary

  initiated by contact of vehicle, e.g. bumper, with an external object, e.g. another vehicle {, or by means of contactless obstacle detectors mounted on the vehicle} · CPC title

• G08G1/167 Primary

  Driving aids for lane monitoring, lane changing, e.g. blind spot detection · CPC title

• G06F18/256

  of results relating to different input data, e.g. multimodal recognition · CPC title

• B60T2201/022

  Collision avoidance systems · CPC title