Method and system for detecting an elevated object situated within a parking facility

Patent metadata

Field	Value
Publication number	US-11080530-B2
Application number	US-201716461953-A
Country	US
Kind code	B2
Filing date	Oct 9, 2017
Priority date	Nov 23, 2016
Publication date	Aug 3, 2021
Grant date	Aug 3, 2021

Abstract

A method for detecting an elevated object situated within a parking facility, using at least two video cameras that are spatially distributed within the parking facility and whose visual ranges overlap in an overlap area. The method encompasses the following: recording particular video images of the overlap area with the aid of the video cameras; analyzing the recorded video images in order to detect an elevated object in the recorded video images, and ascertaining, based on the recorded video images, whether in the detection of an elevated object the detected elevated object is real. A corresponding system, a parking facility, and a computer program are also provided.

First claim

What is claimed is:

1. A method for detecting an elevated object situated within a parking facility, using at least two video cameras that are spatially distributed within the parking facility and whose visual ranges overlap in an overlap area, the method comprising:
a) recording particular video images of the overlap area using the video cameras, wherein the video images of the at least two video cameras are recorded in chronological succession to determine whether or not a detected elevated object is real;
b) analyzing the recorded video images to detect the elevated object in the recorded video images; and
c) ascertaining, based on the recorded video images, whether in the detection of the elevated object the detected elevated object is real, and at which location in the video images, the detected elevated object moves into the particular video image and/or moves out of the particular video image, wherein it is determined that the detected elevated object is real when the detected object moves into the particular video image at an edge of the video image, and/or moves out of the particular video image at the edge of the video image, and it is determined that the detected elevated object is not real when the detected object appears within the video image or disappears within the video image without having crossed the edge of the video image;
wherein for detecting the elevated object in the recorded video images, the following steps are provided according to the analysis according to step b):
rectifying the recorded video images by at least transforming the recorded video images into a bird's-eye view;
comparing the particular rectified video images to one another in order to recognize a difference in the recorded overlap areas; and
detecting the elevated object based on the comparison.

2. The method as recited in claim 1 , wherein step c) includes ascertaining an object speed, the ascertained object speed being compared to a predetermined object speed threshold value, and based on the comparison, determining whether the detected elevated object is real.

3. The method as recited in claim 1 , wherein step c) includes ascertaining a movement of the detected elevated object, it being ascertained whether the movement of the detected elevated object is plausible, and based on the plausibility check, determining whether the detected elevated object is real.

4. The method as recited in claim 3 , wherein it is ascertained whether the ascertained movement is plausible for the classified object.

5. The method as recited in claim 1 , wherein step c) includes classifying the detected elevated object, it being determined, based on the classification, whether the detected elevated object is real.

6. The method as recited in claim 5 , wherein step c) includes ascertaining an object speed, the ascertained object speed being compared to a predetermined object speed threshold value, and based on the comparison, determining whether the detected elevated object is real, and wherein based on the classification, an object-specific speed threshold value is predefined as the predetermined object speed threshold value for the comparison.

7. The method as recited in claim 1 , wherein when the detected elevated object is not real, one or more of the following actions is/are controlled: stopping all driverless motor vehicles traveling within the parking facility, summoning a service person to the video cameras, carrying out a function check of the video cameras, adjusting a particular setpoint trajectory to be driven along by driverless motor vehicles traveling within the parking facility to bypass a section of the parking facility that encompasses the overlap area, blocking a section of the parking facility that encompasses the overlap area, blocking a story of the parking facility that encompasses the overlap area, transmitting an error message to an operator via a communications network.

8. The method as recited in claim 1 , wherein the overlap area encompasses a travel area for motor vehicles.

9. A system for detecting an elevated object situated within a parking facility, the system including at least two video cameras that are spatially distributed within the parking facility and whose visual ranges overlap in an overlap area, the system configured to:
a) record particular video images of the overlap area using the video cameras, wherein the video images of the at least two video cameras are recorded in chronological succession to determine whether or not a detected elevated object is real;
b) analyze the recorded video images to detect the elevated object in the recorded video images; and
c) ascertain, based on the recorded video images, whether in the detection of an elevated object the detected elevated object is real, and at which location in the video images, the detected elevated object moves into the particular video image and/or moves out of the particular video image, wherein it is determined that the detected elevated object is real when the detected object moves into the particular video image at an edge of the video image, and/or moves out of the particular video image at the edge of the video image, and it is determined that the detected elevated object is not real when the detected object appears within the video image or disappears within the video image without having crossed the edge of the video image;
wherein for detecting the elevated object in the recorded video images, the following steps are provided according to the analysis according to step b):
rectifying the recorded video images by at least transforming the recorded video images into a bird's-eye view;
comparing the particular rectified video images to one another in order to recognize a difference in the recorded overlap areas; and
detecting the elevated object based on the comparison.

10. A parking facility that includes a system for detecting an elevated object situated within the parking facility, the system including at least two video cameras that are spatially distributed within the parking facility and whose visual ranges overlap in an overlap area, the system configured to:
a) record particular video images of the overlap area using the video cameras, wherein the video images of the at least two video cameras are recorded in chronological succession to determine whether or not a detected elevated object is real;
b) analyze the recorded video images to detect the elevated object in the recorded video images; and
c) ascertain, based on the recorded video images, whether in the detection of the elevated object the detected elevated object is real, and at which location in the video images, the detected elevated object moves into the particular video image and/or moves out of the particular video image, wherein it is determined that the detected elevated object is real when the detected object moves into the particular video image at an edge of the video image, and/or moves out of the particular video image at the edge of the video image, and it is determined that the detected elevated object is not real when the detected object appears within the video image or disappears within the video image without having crossed the edge of the video image;
wherein for detecting the elevated object in the recorded video images, the following steps are provided according to the analysis according to step b):
rectifying the recorded video images by at least transforming the recorded video images into a bird's-eye view;
comparing the particular rectified video images to one another in order to recognize a difference in the recorded overlap areas; and
detecting the elevated object based on the comparison.

11. A non-transitory computer re

Assignees

• Bosch Gmbh Robert

Inventors

• Hess Felix
• Lehn Andreas
• Nordbruch Stefan

Classifications

• G08G1/04 Primary

  using optical or ultrasonic detectors · CPC title

• G06V20/41

  Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items (segmenting video sequences G06V20/49) · CPC title

• H04N23/90

  Arrangement of cameras or camera modules, e.g. multiple cameras in TV studios or sports stadiums · CPC title

• G06V20/182

  Network patterns, e.g. roads or rivers · CPC title

• G08G1/015

  with provision for distinguishing between two or more types of vehicles, e.g. between motor-cars and cycles · CPC title