Method for creating an environment map for a processing unit
A method for creating an environment map for a processing unit, which moves independently within an environment, wherein a detection device detects environmental features and/or objects within the environment, and wherein position data of the environmental features and/or objects are processed by an evaluation device into an environment map. In order to achieve intuitive user-processing unit interaction, the detected environmental features and/or objects are processed via 3D reconstruction into a photo-realistic, three-dimensional environment card.
1. A method for operating a processing device ( 1 ) used for cleaning a floor surface, which travels automatically within an environment ( 3 ) according to a preprogrammed traveling and processing strategy, comprising the steps of:
detecting with a detection device ( 4 ) environment features ( 5 ) and/or objects ( 6 ) lin the form of object points (A,B) within the environment ( 3 ),
converting position data of the environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B) with an evaluation device into an environment map ( 2 ),
processing the detected environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B) into the environment map to create a photorealistic, three-dimensional environment map ( 2 ) through 3D reconstruction,
recording the environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B) during a movement by the processing device ( 1 ) within the framework of a floor processing operation,
wherein the environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B), are recorded in chronological sequence from differing perspectives of the detection device ( 4 ), wherein the environment map ( 2 ) is reconstructed from the sequential recordings,
wherein the processing device ( 1 ) travels through the environment ( 3 ) along a movement route ( 9 ), wherein the movement route ( 9 ) has a plurality of measuring locations ( 10 ) spaced equidistantly apart, at which the detection device ( 4 ) records the object points (A, B) of the environment ( 3 ),
wherein from a respective two sequential images, the object points (A, B), are then provided with coordinates of a world coordinate system, thereby yielding a point cloud and converting the point cloud into the three-dimensional environment map ( 2 ) as a three-dimensional representation of the environment ( 3 ), which consists of primitive surfaces,
displaying the environment map ( 2 ) on a display ( 7 ) of an external device ( 8 ), and
allocating a name to a room and/or one of the objects ( 6 ) present in the environment map ( 2 ) by a user of the processing device, which name identifies the room, and/or object and is integrated in control commands for controlling a movement of the processing device ( 1 ), wherein the environment map contains the allocated name such that the rooms or objects on the environment map are provided with names, and wherein the user selects a room and/or object to be approached and cleaned by the processing device through a command that states a cleaning instruction and the name of the selected room and/or object.
2. The method according to claim 1 , wherein the environment features ( 5 ) and/or objects ( 6 ) are recorded in chronological sequence from differing perspectives of the detection device ( 4 ), wherein the environment map ( 2 ) is reconstructed from the sequential recordings.
3. The method according claim 1 , wherein the detection device ( 4 ) is a camera, which records images of the environment ( 3 ).
4. The method according to claim 1 , wherein the detected environment features ( 5 ) and/or objects ( 6 ) are processed in the external device ( 8 ) to generate the environment map ( 2 ).
5. The method according to claim 1 , wherein at least one partial area of the displayed environment map ( 2 ) and/or one of the objects ( 6 ) displayed in the environment map ( 2 ) is rotated, magnified and/or shifted.
6. The method according to claim 1 , wherein one of the objects ( 6 ) in the environment map ( 2 ) is segmented and extracted from the environment map ( 2 ), wherein the extracted object ( 6 ) and/or the environment map ( 2 ) reduced by the extracted object ( 6 ) is imported into a data processing program, and processed further by the data processing program.
7. The method according to claim 1 , wherein a current position of the processing device ( 1 ) in the environment map ( 2 ) is displayed.
8. A processing device in the form of a cleaning device for cleaning a floor surface, which travels automatically within an environment according to a preprogrammed traveling and processing strategy, the processing device comprising:
a motor-blower unit for sucking air loaded with suction material into the processing device ( 1 ) through a suction mouth opening, and
a detection device ( 4 ) for detecting environment features ( 5 ) and/or objects ( 6 ) in the form of object points (A,B) within the environment,
wherein the processing device ( 1 ) is designed to use an environment map ( 2 ) containing position data of the environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B) to navigate within the environment,
wherein the detection device ( 4 ) is configured to record the environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B) during a movement of the processing device ( 1 ) within the framework of a floor processing operation,
wherein the environment map ( 2 ) is a photorealistic, three-dimensional environment map ( 2 ), in which the detected environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B) are processed through 3D reconstruction,
wherein the environment features ( 5 ) and/or objects ( 6 ) in the form of the object points (A,B), are recorded in chronological sequence from differing perspectives of the detection device ( 4 ), wherein the environment map ( 2 ) is reconstructed from the sequential recordings,
wherein the processing device ( 1 ) travels through the environment ( 3 ) along a movement route ( 9 ), wherein the movement route ( 9 ) has a plurality of measuring locations ( 10 ) spaced equidistantly apart, at which the detection device ( 4 ) records the object points (A, B) of the environment ( 3 ),
wherein from a respective two sequential images, the object points (A, B), are then provided with coordinates of a world coordinate system, thereby yielding a point cloud and converting the point cloud into the three-dimensional environment map ( 2 ) as a three-dimensional representation of the environment ( 3 ), which consists of primitive surfaces, and
wherein the environment map ( 2 ) is configured to be shown on a display ( 7 ) of an external device ( 8 ), and wherein the environment map is configured such that a user of the processing device ( 1 ) can allocate a name-to a room and/or one of the objects ( 6 ) present in the environment map ( 2 ) such that the environment map contains the allocated name, which name identifies the room, and/or object and is integrated in control commands for controlling a movement of the processing device ( 1 ), such that the rooms or objects on the environment map are provided with names, and wherein the user selects a room and/or object to be approached and cleaned by the processing device through a command that states a cleaning instruction and the name of the selected room and/or object.