Floor-treatment apparatus and navigation system therefor

• US 10,345,821 B2
• Filed: 04/26/2016
• Issued: 07/09/2019
• Est. Priority Date: 05/13/2015
• Status: Expired due to Fees

• Current Assignee: Universitaet Bielefeld
• Original Assignee: Universitaet Bielefeld
• Inventors: Horst, Michael, Moeller, Ralf, Patzelt, Florian
• Primary Examiner(s): Do, Truc M
• Application Number:
  US15/554,822
• Publication Number:
  US 20180052471A1
• Time in Patent Office: 1,169 Days
• Field of Search: None
• CPC Class Codes: A47L 9/28 Installation of the electric equipment, e.g. adaptation or attachment to the suction cleaner; Controlling suction cleaners by electric means
• CPC Class Codes: G01C 21/206 specially adapted for indoor navigation
• CPC Class Codes: G05D 1/0219 ensuring the processing of the whole working surface
• CPC Class Codes: G05D 1/024 in combination with a laser lasers per se H01S
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means
• CPC Class Codes: G05D 1/0248 in combination with a laser lasers per se H01S
• CPC Class Codes: G05D 1/0272 comprising means for registering the travel distance, e.g. revolutions of wheels measuring distance traversed on the ground by vehicles, e.g. using odometers G01C22/00
• CPC Class Codes: G05D 1/0274 using mapping information stored in a memory device navigation using map-matching G01C21/30
• CPC Class Codes: G05D 1/0297 by controlling means in a control room
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G06F 16/5866 using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
• CPC Class Codes: G06T 3/60 Rotation of whole images or parts thereof
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• CPC Class Codes: G06V 2201/07 Target detection
• CPC Class Codes: H04N 7/183 for receiving images from a single remote source

Abstract

A self-driving floor-treatment device navigates by recording images at respective locations along a travel path by a camera the device travels to form a topological map in a database to which a node is added by the floor-treatment device for every location at which an image is recorded and to which an environment images is associated. Each newly recorded environment image with at least one neighboring location in limited proximity of the location belonging to the newly recorded environment image for which a node and an associated environment image have been stored in the map by the same or another floor-treatment device. A new connection is added from the newly recorded node to the associated node of each and every identified neighboring location on the topological map.

Claims

1. A navigation method for a floor-treatment device that is self-driving, the method comprising the steps of:

controlling drives of the floor-treatment device as a function of obstacle information for at least one real object whose obstacle information is detected by sensors on the floor-treatment device;

recording with the floor-treatment device environment images in temporal or spatial intervals at respective locations along a travel path by a camera carried by the floor-treatment device as the floor-treatment device navigates by visual homing;

forming a topological map in a database to which a node is added by the floor-treatment device for every location at which an environment image is recorded and with each of which at least a respective one of the environment images is associated;

forming connections between each node and at least one other node that already exists on the map, was already added by the floor-treatment device, and with which the object information is associated;

determining through a comparison of each newly recorded environment image at respective location for which a new node is being added to the map for all of the environment images in the database that have already been associated with respective nodes whether there is at least one neighboring location in limited proximity of the location belonging to the newly recorded environment image for which a node and an associated environment image have been stored in the map by the same or another floor-treatment device;

if at least one such neighboring location is present, adding a new connection from the newly recorded node to the associated node of each and every identified neighboring location on the topological map; and

for at least one neighboring location to whose node a new connection has been added to the map, calculating obstacle information pertaining to the current location of the newly recorded environment image from the data stored in the map, and using this to control the drives of the floor-treatment device.

2. The method defined in claim 1, wherein the floor-treatment device continues to move at a spacing from locations whose nodes or environment images have connections to travel-distance information in the topological map starting from a node of a recorded neighboring location.

3. The method defined in claim 1, wherein, in the case of a plurality of the floor-treatment devices treating a floor simultaneously, in order to create a single common topological map, the environment images of all floor-treatment devices are stored in

a single common database in a central data-processing system, or

in an individual database on each floor-treatment device, in which case the contents of all individual databases are identical, or each individual database comprises only a portion of the common map.

4. The method defined in claim 3, further comprising the step of:

communicating the environment images wirelessly by the floor-treatment devices to the single common database or to the respective individual databases, each floor-treatment device communicating for this purpose with the central data-processing system or with every other floor-treatment device.

5. The method defined in claim 3, further comprising the step of:

comparing a current environment image of each floor-treatment device in the central data-processing system with all of the environment images stored in the common database, and

sending back the obstacle information determined by local visual homing to the floor-treatment device that supplied the current environment image for comparison.

6. The method defined in claim 3, further comprising the steps of:

effecting the comparison of a current environment image with the environment images stored in the respective individual database individually by each floor-treatment device, and

determining the obstacle information individually as well.

7. The method defined in claim 1, wherein the comparison of a new environment image with all other stored environment images takes place a first step of

determining a group of environment that satisfy a first comparison criterion in a first less stringent and more quickly calculable comparison with the new environment image, and by a subsequent second step of:

identifying environment images from among the group of environment images identified in the first stage that satisfy a second more stringent and more complexly calculated comparison criterion when compared to the newly identified environment image.

8. The method defined in claim 7, wherein, in the first step, groups of several image lines of one line each are averaged for the current environment image and, after a Fourier transform of every averaged line, a spacing of absolute values of Fourier coefficients and corresponding absolute values of the Fourier coefficients of stored environment images is formed, and a group of environment images is selected from the stored images in which distances each fall below a threshold.

9. The method defined in claim 8, wherein, in the second step, a spacing is formed from the current environment image to each image of the selected group, with the calculation of the spacing being repeated several times for two environment images while one of the environment images is rotated by an angular value in an azimuthal direction and a minimum of all spacings being selected, and with the environment images in such locations being selected as neighboring locations whose spacings fall below a threshold.

10. A navigation method as defined in claim 1 for a plurality of the floor-treatment devices, wherein each of the floor-treatment devices carries out the same navigation method that comprises a wall tracking method in which a wall to be tracked is constituted by both a real wall and by a virtual wall formed by nodes in a topological map jointly constructed by all floor-treatment devices.

11. A self-driving floor-treatment device, comprising a device for treating a floor, as well as drives that are controlled by electronics, wherein it is set up to carry out a method defined in claim 1.