DE102010029241A1 - Soil cultivation device and method for processing a floor surface - Google Patents

Abstract

The invention relates to a method for processing a soil surface by means of a soil cultivation device, which has a soil cultivation unit and at least one detection unit, the soil cultivation device being moved over the floor area, with the at least one detection unit being used to create images of the soil surface at successive times and from the soil cultivation device using the Images a map of the floor area is generated. In order to provide such a method, in which a map of the floor area is generated with relatively little equipment and computational effort as a basis for the most comprehensive processing of the same, it is proposed that the images from the floor cultivation device are examined to determine whether the floor area is a regularly repeating one Has basic pattern, a ground surface segment having the basic pattern is identified by the soil cultivation device as a ground surface segment to be processed and that an area assigned to this ground surface segment is stored in the map for each ground surface segment to be processed. The invention also relates to a soil cultivation device for carrying out the method.

Description

The invention relates to a method for processing a ground surface by means of a harrow having a soil treatment unit and at least one detection unit, wherein the harrow is moved over the ground surface, by means of the at least one detection unit at successive times images of the soil surface are created and from the harrow on the basis of Images a map of the floor surface is generated.

The invention also relates to a harrow for performing the method with a soil treatment unit and with at least one detection unit for creating images of the soil surface to be processed at successive times during movement of the harrow, wherein the floor cleaning device based on the images a map of the soil surface can be generated.

The generation of a map of the ground surface to be processed is known in self-propelled and self-steering tillage equipment. These allow the processing, such as cleaning, a floor area with the help of the card without the use of an operator. For example, in the EP 1 768 008 A1 described such a working according to a generic method tillage device of the type mentioned. This harrow creates optical images of the ground surface at very high repetition rates to generate the map and for self-localization. However, this requires a considerable computational effort to create the map with sufficient accuracy and to determine the position of the harrow on the ground surface with sufficient accuracy. However, this is necessary in order to cover the floor area as comprehensively as possible.

An efficient area-wide processing of the ground surface in the shortest possible time is always desirable for achieving a good cost-benefit ratio, be it in an operation of a self-propelled and self-steering harrow or even with the use of an operator for the harrow.

In addition, when processing the floor surface, it should also be considered that the floor surface may be divided into areas to be worked on and not to be machined, and for this reason only a partial processing of the floor surface is required, possibly even only permitted.

Object of the present invention is to produce a map of the bottom surface as a basis for the widest possible coverage of the same with a relatively low equipment and computational effort.

This object is achieved in a generic method according to the invention that the images are examined by the harrow to see if the bottom surface has a regularly repeating basic pattern, the ground pattern exhibiting bottom surface segment of the harrow is identified as to be processed bottom surface segment and in the map for each Processing floor area segment is stored an area associated with this floor area segment.

The idea flows into the invention that, in certain cases, the floor area to be worked on has a pattern composed of a regularly repeating basic pattern. The bottom surface may therefore comprise at least sections regularly arranged and each having the basic pattern having bottom surface segments. For example, a tiled bottom surface is configured such that each bottom surface segment is formed by a tile, which usually defines a rectangular basic pattern. Examples of buildings with such trained floor areas are buildings with a lot of public traffic, such as supermarkets, office buildings, public buildings, etc. Here, it is often necessary to edit the tiled areas of the floor surface and in particular to clean, whereas any additional not Tiled areas, which are lined with carpet, do not need to be cleaned or may not be cleaned.

According to the invention, therefore, the method provides for analyzing the images in such a way as to determine whether the bottom surface is composed of a base pattern having regularly repeating bottom surface segments. In addition, it is assumed that a bottom surface segment having the basic pattern is to be processed. On the basis of this information, a map of the floor area can be created in a simple manner, for example by a control unit of the soil cultivation device, in which the actual floor surface segments to be processed are assigned to individual areas within the map. The analysis of the images and the creation of the map can be carried out with relatively little equipment and computational effort, as a rule, since the tillage device can transmit a pre-existing and in practice often simple constructed pattern of the bottom surface in a sense in the card.

Based on the map, an efficient processing of the ground surface can be carried out. This is possible in the case of control of the processing device by an operator who can recognize, control and work with the soil processing unit on the basis of the map to be processed bottom surface segments. Even with an autonomous processing of the ground surface by a self-propelled and self-steering tillage implement can be run over as recognized for processing soil surface segments and processed by means of the soil cultivation unit.

The map of the bottom surface can be easily generated when it is divided into cells, with each bottom surface segment having the basic pattern being associated with a cell. A map created in this way can be managed and processed with little computational effort and managed, for example, in the form of a matrix.

It can be provided that even the basic pattern not having bottom surface segments in the map, for example in the cells in the last-described variant of the method, are registered and characterized as "not to edit".

It is advantageous if features of the images are extracted and comparison patterns constructed from the features are compared for agreement with one or more reference patterns stored in a memory element. The feature extraction, the construction of comparison patterns and their comparison with one or more reference patterns can be performed with low computing capacity. This makes it possible to determine with little expenditure on equipment, whether the floor surface has a regularly repeating basic pattern.

It is advantageous if edges or lines are detected as features and / or if quadrilaterals are compared as comparison patterns. This allows a reliable and at the same time with comparatively low computing power feasible determination of whether the bottom surface has a regularly repeating basic pattern. In addition, the practice is taken into account that patterned to be processed floor surfaces are often tiled and consequently typically have rectangular basic pattern. The reference patterns in this and the previous variant of the method are also preferably quadrangular and in particular rectangular.

It can be provided that other polygons than quadrangles, such as triangles, are compared as a comparison pattern. It is also possible to compare composite comparison patterns from a plurality of polygons, for example a plurality of rectangles.

It can be conveniently further provided that based on extracted features, a new pattern is created and stored as a reference pattern in the memory element, so that it can be considered in the future analysis of the images and used to examine the floor surface on a regularly repeating basic pattern.

Preferably, by comparing two or more successive images, the position of the harrow relative to the bottom surface segments having the basic pattern is determined as a function of time. The movement of the harrow on the ground surface causes two or more successive images to detect different portions of the ground surface. In this variant of the method is therefore made possible to determine a change in position of the harrow. The position can be stored time-dependent in the map. This can be achieved with little computational effort, a self-localization of the harrow and further its movement path can be determined.

It is advantageous if images of the floor area are created at regularly successive times. If the tillage implement, as usual, moves over the ground surface at a constant speed, this facilitates the determination of the position and the determination of the travel path of the tillage implement.

Preferably, the times at which images of the bottom surface are created, determined depending on the speed of the harrow. This can ensure that the accuracy of the comparison of successive images is sufficiently large in order to achieve a reliable position determination and determination of the movement path of the harrow.

In one embodiment of the method in practice, it is favorable for a sufficiently accurate position determination and determination of the movement path when the harrow at a speed of about 0.2 to about 1.5 m / s and preferably from about 0.4 to 1 m / s is moved. It will be conveniently created one to about 10 images per second and preferably 4 to 6 images per second. These can be computationally easily analyzed for the presence of regular basic patterns.

Advantageously, as stored in the map, for example, when the harrow moves over a bottom surface segment having the base pattern and is processed. This allows efficient processing of the floor area. Already processed ground surface segments may be marked as such and, if avoidable, need not be run over again and processed.

Preferably, an operator of an image of the map is displayed by means of a display unit. Information stored in the card is thereby recognizable to the operator to assist in controlling the harrow. In the last-described variant of the method, the operator z. B. recognize which the basic pattern having ground surface segments have already been run over and edited and which are still to be processed. This allows efficient processing of the floor surface, because the operator can target and edit unprocessed floor surface segments targeted. At the same time a multiple processing of floor surface segments can be largely avoided.

With manual control of the harrow, the operator can ensure that direction changes are made in a suitable manner, so that, for example, the ground pattern not having ground surface segments are not run over for their protection and / or not processed.

It is advantageous if the ground surface is processed autonomously by means of the cultivator, the cultivator being self-propelled and self-steering and having a drive unit and a control unit for controlling the movement of the floor-cleaning appliance. This allows the cost-effective processing of the floor surface without the use of an operator. As to recognized to edit floor surface segments can be processed autonomously by the control unit drives the drive unit so to drive over the bottom surface segments and to edit by means of the soil cultivation unit. This can be ensured in a simple manner and with low equipment and computational effort that the widest possible coverage of the actual to be processed segments floor area is achieved.

Preferably, images of portions of the ground surface ahead of the cultivator are created, based on its direction of travel based on a locomotion pattern. Thereby, the direction of movement of the harrow can be maintained when it is determined that the ground surface ahead has the regularly repeating basic pattern. As a result, a more rapid processing of the soil can be achieved.

The harrow can thus move over the ground surface based on one or more locomotion patterns. It is advantageous if, in accordance with the basic locomotion pattern of the cultivator, an in particular straight-ahead travel is carried out, since this is a particularly simple locomotion pattern. In combination with the last-described variant of the method, the harrow can therefore drive straight ahead until it is determined that the area of the ground surface lying ahead no longer has the regularly repeating basic pattern.

Preferably, the state of movement of the harrow is changed when it is determined that in the direction of movement of the harrow no Grundflächen having ground surface segments are no longer present. In this way it can be achieved, for example, that the bottom surface segments not having the basic pattern are not run over and processed. This is advantageous, for example, if there are floor surface segments that can not be edited.

For example, as a change of the state of motion, the harrow can be stopped and / or a change in the direction of movement of the harrow can be performed. For example, it is possible for the harrow to rotate on the ground surface to determine if the ground surface imaged after rotation has the regular repeating pattern so that further processing of the ground surface can proceed.

It is advantageous if, after the change of the state of motion of the harrow, the drive unit is controlled by the control unit so that not yet processed, but an area of the map already assigned and the basic pattern having bottom surface segments are run over and edited. As a result, the area-wide processing of the floor surface can be ensured.

In general, it is favorable if the drive unit is controlled by the control unit such that the basic pattern of non-exhibiting bottom surface segments is not run over and / or not processed. As a result, it is possible that the bottom surface segments marked as "not machined" are spared. In the example mentioned at the beginning of the partly tiled and partly with Carpet-lined floor surface, it is possible, for example, that only the tiles may be cleaned, but the carpet must not be run over or cleaned.

In the method according to the invention, the size of the bottom surface or the bottom surface segments having the basic pattern need not be known. Nevertheless, it is favorable if the size of the base surface segments having the basic pattern is determined on the basis of the images. This allows, for example, by comparing successive images to determine the distance covered by the tillage implement.

It is advantageous if, by means of at least one Radencoders, the distance covered by the soil cultivation device and / or changes in direction of the soil cultivation device are recorded and, in particular, stored time-dependently. This allows to determine the movement path of the tillage implement.

Preferably, based on the images provided by the at least one Radencoder encoder data based on the images are checked for plausibility. This allows, for example, a more reliable determination of the position and the movement path of the harrow. For this purpose, the distance determined on the basis of the encoder data can be compared with the distance which is determined on the basis of the evaluation of two or more successive images. Also, the size of the basic pattern having bottom surface segments, which are determined based on the images can be checked by means of the encoder data.

In general, the principle is that the greater the time between two time points at which images of the bottom surface are created, the greater the accuracy of the encoder data, and vice versa. It can be provided that in the case of a negative plausibility check of encoder data with data determined on the basis of the images, the rate of creating images is increased. Additionally or alternatively, the speed of the harrow can be reduced.

It has already been mentioned that it is advantageous if in the map for the basic pattern not having ground surface segments each one of these floor space segments associated area is stored, so that these floor space segments are deposited in the map and the card is completed.

For the same reason, it is favorable if, with one or more sensors, obstacles on the floor surface and / or boundaries of the floor surface are detected and stored as areas assigned to the obstacles or as boundaries of the floor area in the map. This will complete the map. A self-propelled and self-steering tillage implement can use this information to determine where the ground surface is unmanageable and, in particular, to avoid collisions with obstacles or boundaries on the ground surface.

The inventive method is particularly suitable for processing a floor surface by means of a floor cleaning device, in particular a self-propelled and self-steering floor cleaning device.

Object of the present invention is also to provide a harrow of the type mentioned for performing the method.

This object is achieved in a generic harrow according to the invention that the tillage equipment can determine whether the bottom surface has a regularly repeating basic pattern that the basic pattern exhibiting bottom surface segment of the harrow is identified as being processed and that the harrow comprises a storage member for storing a a region of the map assigned to a surface area segment to be processed.

Such an embodiment of the soil cultivation device makes it possible with comparatively little apparatus and computational effort to create a map of the floor surface as a basis for the widest possible coverage of the same.

It is advantageous if the at least one detection unit is designed as an optical detection unit, in particular as a digital camera. This allows a structurally simple creation of images of the floor surface. In order for these to have a good contrast, the detection unit preferably makes black and white images of the bottom surface, for example with a resolution of 640 × 480 pixels to approximately 2 megapixels.

It is advantageous if the soil cultivating device comprises a pattern recognition element for extracting features of the images and for constructing comparative patterns, and a comparison element for comparing the comparison patterns with one or more reference patterns stored in the memory element. This makes it possible, with relatively little outlay on equipment and computation, to reliably determine whether the floor surface to be machined has a regularly repeating basic pattern.

Preferably, the tillage implement comprises a locating member for determining the position of the tillage implement relative to the ground plane segments having the base pattern on the basis of comparative data of the comparison member obtainable by comparing two or more consecutive images of the ground surface. Thereby, the position of the harrow on the ground surface can be determined and stored. By time-dependent storage of the position changes in position and the movement path of the harrow can be tracked.

It is expedient if the cultivator comprises in each case a wheel encoder associated with a drive wheel for determining the number of revolutions of the respective drive wheel. As a result, the distance covered by the cultivator and changes in direction along the path can be tracked. On the basis of the encoder data, for example, the position of the soil cultivation device, which is determined from based on the images data are checked.

Preferably, the soil cultivation device comprises a computing element for determining the size of a base surface segment having the basic pattern and / or for determining the distance covered by the soil cultivating device. The size of the bottom surface segment with basic pattern can be calculated because the field of view of at least one detection unit and its position on the harrow is known. By comparing two or more successive images of the ground surface, the computing element can also determine the path of movement of the soil cultivation device. Furthermore, encoder data can be used to determine the route and to determine the size of the floor surface segment.

Advantageously, the harrow comprises one or more sensors for detecting obstacles on the ground surface and / or for detecting boundaries of the ground surface. As a result, additional information characterizing the surface to be processed can be obtained and entered in the map. This will complete the map. This facilitates the control of the movement of a self-propelled and self-steering tillage implement over the ground surface to be processed.

For detecting obstacles and / or boundaries of the ground surface, the cultivator may, for example, comprise a distance sensor, such as an infrared sensor, an ultrasonic sensor, a radar sensor or a laser sensor. It is also possible to use a touch-sensitive sensor which detects the obstacles and / or boundaries of the floor surface in a groping manner.

The tillage implement preferably comprises a display unit for displaying an image of the map. The image of the map may assist an operator controlling the tillage implement, for example by marking already processed and not yet processed ground plane segments with basic patterns.

It is favorable if the harrow is designed to be self-propelled and self-steering and has a drive unit and a control unit for controlling the movement of the harrow. This allows in the manner explained above, the autonomous processing of the floor surface.

The harrow is preferably designed as a floor cleaning device. The soil working unit may in this case comprise, for example, a scrubbing unit and / or a dirt receiving unit. Additionally or alternatively, for example, a suction unit and / or a polishing unit may be provided.

The following description of a preferred embodiment of the invention serves in conjunction with the drawings, the detailed explanation of the invention. Show it:

1 a perspective view of a soil cultivation device according to the invention on a floor surface to be processed;

2 a block diagram of the control system of the tillage implement;

3 a plan view of a section of a ground surface to be processed and the harrow at two different times in a schematic representation;

4 : A schematic representation of a section of a map created by the soil cultivating device of the bottom surface 3 at a first time;

5 : in the 4 displayed map at a later time;

6 FIG. 2: a plan view of an operating unit of the soil cultivating device, comprising a display unit and

7 : an enlarged view of the display unit of the control unit off 6 ,

In 1 is shown in perspective a preferred embodiment of a harrow according to the invention, the control system in 2 block-like shown. The harrow is denoted by the reference numeral 10 occupied and designed as a mobile, self-propelled and self-steering floor cleaning device. It enables the autonomous cleaning of a floor surface 12 which is believed to have, at least in sections, a regularly repeating basic pattern. This will be explained in more detail below.

To move on the floor surface 12 includes the floor cleaning device 10 a suspension 14 with two drive wheels rotatable about a common axis of rotation 15 and 16 which drive motors 18 respectively. 19 assigned. The drive motors 18 and 19 are part of the chassis 14 and they are over control lines 21 respectively. 22 with a control unit 24 as well as with known, not shown in the drawing electric batteries in electrical connection. The drive wheels 15 . 16 and the drive motors 18 . 19 form a drive unit 25 of the floor cleaning device 10 for automatically moving over the floor surface 12 ,

In addition, the floor cleaning device 10 be controlled by one operator by one at the back 26 of the floor cleaning device 10 arranged operating unit 28 served.

The floor cleaning device 10 is as a scrubber 10 designed and has a scrubbing unit 30 on, the lower side at the front 32 of the floor cleaning device 10 is arranged, and a dirt receiving unit 34 at the back 26 , By means of the scrubbing unit 30 can remove dirt from the floor surface 12 using one in a tank 36 stored cleaning liquid are replaced. The mixture of cleaning fluid and dirt can by means of the dirt receiving unit 34 from the bottom surface 12 be taken and in a dirty liquid container 38 be transferred. The scrub unit 30 and the dirt pickup unit 34 form one in the 2 with the reference number 40 occupied tillage and in particular floor cleaning unit of the floor cleaning device 10 , The floor cleaning unit 40 stands with the control unit 24 via a control line 42 in active connection.

Near the front 32 includes the floor cleaning device 10 a detection unit 44 in the present case is an optical detection unit and in particular a digital camera 46 is. The registration unit 44 stands with the control unit 24 via a signal line 48 in active connection. The digital camera 46 is on an approximately vertically extending support 50 at a height of about 0.5 to about 1.5 meters and preferably about one meter above the ground surface 12 at the floor cleaning device 10 assembled. She looks at an angle with a field of view 52 on one in front of the floor cleaning device 10 located section of the floor area 12 , so by means of the digital camera 46 Images of the floor area 12 can be created.

The digital camera 46 has a resolution of about 640 × 480 pixels to about 2 megapixels. This ensures that a sufficient resolution of on the bottom surface 12 existing structures made possible by the digital camera 46 created images of the floor area 12 but still have only a small amount of data.

By means of the digital camera 46 are images of the floor surface at successive intervals, preferably regularly following each other 12 created. During the movement of the floor cleaning device 10 For example, from about 0.2 to about 1.5 m / s, for example, from 1 to about 10 images per second can be made. The images can be sent to the control unit 24 over the signal line 48 to be provided.

The control unit 24 includes a pattern recognition element 54 for extracting features contained in the images. For example, the pattern recognition element 54 Detect edges or lines in the images and create comparison patterns from these detected edges or lines. The comparison patterns, such as rectangular patterns, can be detected by means of one of the control unit 24 included comparator 56 be checked for compliance with one or more reference patterns stored in a memory 58 the control unit 24 are stored. Represents the comparison member 56 notice that one of the pattern recognition member 54 provided comparison pattern a predetermined minimum degree of compliance with a reference pattern in the memory element 58 has, can thus on the basis of the images of the bottom surface 12 to determine if the floor area 12 has a basic pattern matching the reference pattern. In particular, it can be determined whether the bottom surface has a regularly repeating basic pattern.

This is the case in the case of the floor cleaning device 10 especially because of importance, because the floor cleaning device 10 used for wet scrubbing of tiled floors. Accordingly, the floor area 12 in the present example for explaining the invention partially from a plurality of identical square tiles 60 educated. The square Surface of each tile 60 defines a square basic pattern, which is in the area where the floor area 12 tiled, repeated regularly. Based on the images of the floor surface 12 can thus be determined which sections of the floor area 12 with tiles 60 are occupied. This finding allows easy generation of a map of the floor surface to be cleaned 12 , and the map can be easily a nationwide cleaning of the floor surface, as far as it is to be performed, can be achieved. This will be explained below with reference to the 3 to 5 explained.

3 schematically shows a plan view of the bottom surface 12 which one with tiles 60 tiled main corridor 62 and a side corridor branching off from it at right angles 64 also with tiles 60 tiled. The side corridor 64 Opposite borders the main corridor 62 It surrounds it from two sides and with a carpet floor 66 lined floor area 68 , It is assumed that the instruction exists that the floor cleaning device 10 the tiles 60 should clean the carpet 66 but neither overrun nor clean may.

Such, with tiles 60 partially tiled floor space exists for example in buildings with a lot of public traffic, for example in office buildings, public buildings, supermarkets etc.

At a first time is the floor cleaning device 10 in the main corridor 62 and moves along a movement pattern with an arrow 70 illustrated movement direction. The movement pattern of the floor cleaning device 10 is usually such that it moves in straight straight ahead. This will be the drive motors 18 and 19 from the control unit 24 so controlled that the drive wheels 15 and 16 turn evenly. At first, the field of view captures 52 the digital camera 46 a section of the main corridor 62 and a portion of the floor area 68 ,

Based on the pattern recognition element 54 and the comparison element 56 puts the control unit 24 determines that part of the image captured by the ground surface 12 has a regularly repeating basic pattern. The basic pattern of the floor surface segments are called tiles 60 identified with a square basic pattern. In the present case, these are the tiles A-0 to A-2, B-0 to B-2 and C-1 and C-2, wherein the reference numbers are only illustrative using a coordinate system 71 of the 3 serve. An area of the floor area adjacent to the tiles A-0 to A-2 68 with carpet 66 is recognized as a bottom surface segment without the basic pattern.

As in the floor cleaning device 10 is saved, that only ground surface segments with regular basic pattern, but not bottom surface segments are to be cleaned without the basic pattern, a map 72 the floor area 12 are created, are distinguished in the cleaned floor surface segments of non-cleanable floor surface segments. The map 72 is in the memory element 58 as a matrix 74 of sufficiently large extent and is in individual cells 76 divided, the size and shape of which can be selected depending on the shape of the recognized basic pattern. This means that the cells 76 be chosen square. Each floor surface segment as a tile 60 has been recognized, is in the card 72 entered as to be cleaned floor area segment. Therefore, at the first time, cells A-0 to A-2, B-0 to B-2, and C-1 and C-2 in the card become 72 Cells a-0 to a-2, b-0 to b-2 and c-1 and c-2 assigned in a unique manner. These cells are in 4 marked by a dash.

Even those floor surface segments that are not recognized as having the basic pattern will be on the map 72 entered. They too become cells 76 assigned, which are marked as "not clean". This is present in 4 for the cells z-1 and z-2 shown in the field of view 52 covered area of the ground area 68 with carpet 66 as a non-machinable area of the floor area 12 is stored. These cells are in 4 marked by a cross.

While the floor cleaning device 10 along the direction of movement 70 moved, images of the floor surface are now periodically 12 created. Gradually, floor surface segments become the floor surface 12 showing the square basic pattern of the tiles 60 show, the card 72 added. Thereby every tile becomes 60 a cell 76 assigned and marked as "to clean". Similarly, during the movement more than "not to be cleaned" bottom surface segments, as not with tiles 60 tiled, in the map 72 marked as "not to be cleaned". In this way, the map gradually emerges 72 the floor area 12 ,

By comparing two or more successive images of the floor surface 12 can by the pattern recognition element 54 be noted that a movement of the floor cleaning device 10 relative to the floor area 12 is done. This recognizes the pattern recognition element 54 Remember that in the images the tiles 60 appear in consecutive images at different positions. By means of a localization member 78 the control unit 24 This can change the position of the floor cleaning device 10 on the floor surface 12 relative to the tiles 60 Tracked time-dependent and, for example, also in the map 72 get saved. This allows the position of the floor cleaning device 10 on the previously mapped floor area 12 determined and also the path of movement of the floor cleaning device 10 over the floor area 12 determined and possibly also in the map 72 be registered.

On the basis of a computing element 80 the control unit 24 can change the size of the tiles 60 be calculated and based on this information further from the floor cleaning device 10 distance traveled. This is possible because the position of the digital camera 46 at the floor cleaning device 10 and the size of the field of view 52 the digital camera 46 is known.

The from the floor cleaning device 10 Distance traveled can also be determined by the number of revolutions of the drive wheels 15 and 16 is detected. For this purpose, the floor cleaning device 10 the drive wheels 15 and 16 assigned Radencoder 82 respectively. 83 on that with the control unit 24 via signal lines 85 respectively. 86 are connected. This makes it possible to use the images of the floor surface 12 through the calculator 80 Check the distance traveled for plausibility. This allows the position determination of the floor cleaning device 10 relative to the floor area 12 , which is determined on the basis of their images, checked and, if necessary, corrected. The result is a more accurate self-localization of the floor cleaning device 10 ,

Because the position of the floor cleaning device 10 can be determined in the manner described above, it is possible to determine which of the already recognized tiles 60 run over and cleaned. The run over and cleaned tiles 60 be in the map 72 marked as cleaned. This is in 5 by a plus sign in the corresponding cell 76 symbolizes.

As is the floor cleaning device 10 By default, it moves in a straight line straight ahead, it drives in the present case to a limit 88 at the end of the main corridor 62 at the bottom of this area 68 with carpet 66 borders. Once it is there, the tiles A-0 to A- (n + 2) and B-0 to B- (n + 2) are recognized and already cleaned, and the tiles C-1 to C- (n + 2) are recognized, but not yet cleaned. The along the tile row A extending portions of the floor area 68 with carpet 66 were recognized as bottom surface segments not having the basic pattern, and they are in the column z in the map 72 saved.

On the border 88 arrived, causes the control unit 24 the drive motors 18 and 19 to stop, as the floor cleaning device 10 lying ahead only floor surface segments of the floor area 68 without the square basic pattern being recognized. This means that the state of motion of the floor cleaning device 10 changes, and the control unit 24 then causes a rotation of the floor cleaning device 10 by appropriate control of the drive motors 18 and 19 ,

During rotation, the floor cleaning device detects 10 further tiles arranged in rows of tiles n + 1 and n + 2 60 with square basic pattern and along the border 88 lying floor surface segments of the floor area 68 with carpet 66 without square basic pattern and carries these in each case in the map 72 one.

In the further course, the control unit controls 24 the drive motors 18 and 19 such that the already recognized but not yet cleaned tiles 60 the row of tiles C are cleaned. The floor cleaning device 10 will therefore be parallel to the original direction of movement 70 along a by an arrow 90 symbolized movement direction again in straight-line straight ahead on the rows of tiles C and D moves. Also in this case, the position of the floor cleaning device is continuously 10 based on the images of the floor area 12 determined and stored time-dependent and based on the encoder data of the Radencoder 82 and 83 controlled.

Newly recognized nasty 60 be in the card 72 registered (cross-hatch in the 4 and 5 ), and recognized and already cleaned tiles are in the map 72 marked as such (plus sign in the 4 and 5 ). At a later date, to which the contours of the floor cleaning device 10 in 3 are shown by dashed lines and this by the reference numeral 10 ' is occupied, takes the card 72 therefore the in the 5 represented shape. At this later time, the tiles are A-0 to A- (n + 2), B-0 to B- (n + 2), C-2 to C- (n + 2), and D-2 to D- ( n + 2), and the tiles C-0, C-1, D-0, D-1, E-0 to E- (n + 2) and Fn to F- (n + 2) are already recognized, but not cleaned yet.

In the further course of the cleaning process sets the floor cleaning device 10 First, the cleaning of the main corridor 62 continues and cleans tile rows C and D.

In the manner described above, successive rows of tiles A to F are to be cleaned with tiles 60 occupied detected and these tiles 60 be cleaned. Moves the floor cleaning device 10 along the tile rows E and F, are in the side corridor 64 existing tiles 60 recognized with basic pattern and in the map 72 entered. Is therefore the main corridor 62 Completely cleaned, sets the floor cleaning device 10 the cleaning of the floor area 12 in the side corridor 64 continued. Do not become tiles 60 more recognized with a square basic pattern, that is the proportion of the floor area 12 that with tiles 60 tiled and therefore is to be cleaned, completely cleaned area-wide. During the cleaning process has the floor cleaning device 10 the map 72 of the area of the floor surface to be cleaned 12 self made. The map 72 For example, for future applications in the memory member 58 get saved.

As an additional tool to create the map 72 and to move over the floor surface 12 includes the floor cleaning device 10 a sensor unit 92 connected to the control unit 24 via a signal line 94 connected is. The sensor unit 92 includes a sensor element 96 and a sensor element 98 on a side surface or on the front side 32 of the floor cleaning device 10 on. At the sensor elements 96 and 98 are distance sensors, z. In the form of infrared sensors, ultrasonic sensors, radar sensors, laser sensors or the like. This allows the floor area 12 to capture comprehensive limitations and this information to the control unit 24 provide. For example, the floor area becomes 12 adjacent to the row of tiles F of walls 99 and 100 limited. The wall 99 can from the sensor elements 96 and 98 be recognized as such when the floor cleaning device 10 moves over the rows of tiles A and B This allows it in the map 72 characterize cells g-4 to g- (n + 2) immediately as "non-clean" areas. A separate marking that these are not floor area segments but boundaries of the floor area 12 can also be made.

It is also possible, based on the sensor unit 92 the presence of obstacles also the floor area 12 to capture and in the map 72 as "not to clean" to enter. By suitable control of the drive motors 18 and 19 through the control unit 24 These obstacles may come from the floor cleaning device 10 be driven around.

The floor cleaning device 10 can except in the above-described self-propelled and self-steering operating mode, in which the bottom surface 12 is autonomously cleaned, operated in two other modes of operation, in which an operator of the floor cleaning device 10 using the control unit 28 controls. In a first, manual operating mode of the two other operating modes, the operator controls the floor cleaning device 10 sliding and directing by means of two handle elements 102 and 104 the operating device 28 ( 6 ). In a third, semi-automatic operating mode, the operator can use the floor cleaning device 10 by means of the handle elements 102 and 104 to steer. In addition, it can by means of an actuating lever 106 the drive unit 25 activate so that the floor cleaning device 10 Effortless over the floor surface 12 can be moved.

The operating unit 28 further includes another operating lever 108 for controlling the floor cleaning unit 40 , a display unit 110 and a plurality of controls 112 ,

Operating instructions of the operator on the operating unit 28 can the control unit 24 via a control line 114 be transmitted. The control of the display unit 110 through the control unit 24 via a further control line 116 ( 2 ).

The display unit 110 serves on the display of the operating state of the floor cleaning device 10 information in addition to a picture 118 the map 72 display. This proves to be in the semi-automatic and in the manual operating mode of the floor cleaning device 10 as very helpful to the operator, because in the map 72 Information contained by means of the display unit 110 be visualized. This can be z. B. also in the form of a matrix 120 take place, which is constructed identically to the matrix 74 , Accordingly, the image is 118 the map 72 in a variety of cells 122 divided, each cell 76 assigned.

In 7 is the image 118 the map 72 presented at a time when the floor cleaning device 10 by the operator through the tile rows C and D along the direction of movement 90 is moved. A contour 124 that in the image 118 is displayed, the operator signals the current position of the floor cleaning device 10 ,

The visualization of the map 72 allows the operator to determine which of the basic patterns having tiles 60 the floor cleaning device 10 has already been moved away during the cleaning process and thus, which of the tiles 60 already cleaned. These tiles 60 are for the operator in the image 118 by the plus sign as in the map 72 characterized. In a similar way, the operator can recognize which tiles 60 Although recognized as such, they have not yet been purified. These tiles 60 are in the image 118 like in the map 72 marked by a dash.

On the display unit 110 superimposed touch-sensitive navigation elements 126 serve to move the illustrated section of the image 118 the map 72 so that the Operator a better overview of the image 118 can provide.

The visualization of the map 72 makes it easier for the operator, the floor cleaning device 10 to control that already cleaned areas of the floor area 12 not to be cleaned again and not yet cleaned areas of the floor area 12 be targeted and cleaned. This enables efficient, area-wide cleaning of the floor area 12 ,

The cleaning assistant function of the floor cleaning appliance described above 10 can also be used in a floor cleaning device, which in contrast to the floor cleaning device 10 is not self-propelled and / or self-steering, but otherwise the features of the floor cleaning device 10 having. By means of such a floor cleaning device which can be operated only in a manual and / or in a semi-automatic operating mode, an efficient area-wide cleaning of the floor surface can likewise be achieved by utilizing the cleaning assistance function.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• EP 1768008 A1 [0003]

Claims (30)

A method for processing a ground surface by means of a harrow having a soil treatment unit and at least one detection unit, wherein the tillage device is moved over the bottom surface, images of the bottom surface are created by means of the at least one detection unit at successive time points and the map of the soil cultivation device based on the images Floor surface is generated, characterized in that the images are examined by the tillage device to see whether the bottom surface has a regularly repeating basic pattern, a Grundflächen having Bodenflächensegment is identified by the harrow as a Bodenflächensegment to be processed and in the map for each to be processed bottom surface segment a this Floor area segment associated area is stored. A method according to claim 1, characterized in that features of the images are extracted and comparison patterns constructed from the features are compared for agreement with one or more reference patterns stored in a memory element. Method according to claim 2, characterized in that edges are detected as features and / or that quadrilaterals are compared as comparison pattern. Method according to one of the preceding claims, characterized in that by comparing two or more successive images, the position of the harrow relative to the base pattern having bottom surface segments is determined time-dependent. Method according to one of the preceding claims, characterized in that images of the ground surface are created at regularly successive times. Method according to one of the preceding claims, characterized in that the times at which images of the ground surface are created, are determined depending on the speed of the soil cultivation device. Method according to one of the preceding claims, characterized in that it is stored when the harrow moves over a base pattern segment exhibiting the base pattern and this is processed. Method according to one of the preceding claims, characterized in that an operator of an image of the card is displayed by means of a display unit. Method according to one of the preceding claims, characterized in that the bottom surface is processed autonomously by means of the harrow, wherein the harrow is self-propelled and self-steering and has a drive unit and a control unit for controlling the movement of the harrow. A method according to claim 9, characterized in that images of the tillage device ahead of lying portions of the bottom surface are created, based on the direction of movement based on a locomotion pattern. A method according to claim 9 or 10, characterized in that after basic locomotion pattern of the harrow straight ahead driving is performed. Method according to one of claims 9 to 11, characterized in that the movement state of the harrow is changed, if it is determined that in the direction of movement of the harrow no basic pattern having ground surface segments are no longer present. A method according to claim 12, characterized in that the soil cultivation device is stopped and / or that a change in the direction of movement of the soil cultivation device is performed. Method according to one of claims 9 to 13, characterized in that after the change of the state of motion of the harrow, the drive unit is controlled by the control unit so that not yet processed, but an area of the card already assigned and the basic pattern having bottom surface segments are run over and edited , Method according to one of claims 9 to 14, characterized in that the drive unit is controlled by the control unit such that the basic pattern not having Bodenflächensegmente not run over and / or not processed. Method according to one of the preceding claims, characterized in that based on the images of the bottom surface, the size of the basic pattern having bottom surface segments and / or the distance covered by the tillage device is determined. Method according to one of the preceding claims, characterized in that by means of at least one Radencoders covered by the cultivator distance and / or changes in direction of the soil cultivation device are detected and stored in particular time-dependent. A method according to claim 17, characterized in that based on the data provided by the at least one Radencoder provided encoder data based on the images are checked for plausibility. Method according to one of the preceding claims, characterized in that in the map for the basic pattern not having ground surface segments each one of these floor surface segments associated area is stored. Method according to one of the preceding claims, characterized in that with one or more sensors obstacles on the bottom surface and / or boundaries of the bottom surface are detected and stored as the obstacles associated areas or the boundaries of the bottom surface in the map. Soil cultivation device ( 10 ) for carrying out the method according to one of the preceding claims with a soil treatment unit ( 40 ) and at least one registration unit ( 44 ) for creating images of the ground surface to be processed ( 12 ) at successive times during the movement of the harrow ( 10 ), whereby the tillage equipment ( 10 ) using the images a map ( 72 ) of the floor surface ( 12 ) is producible, characterized in that the tillage device ( 10 ) it is possible to determine whether the floor area ( 12 ) has a regularly repeating basic pattern such that a bottom surface segment comprising the basic pattern is separated from the soil cultivating device ( 10 ) is identifiable as being editable and that the soil tillage implement ( 10 ) a memory member ( 58 ) comprises for storing a region in the map assigned to a surface area segment to be processed ( 72 ). Soil cultivation device according to claim 21, characterized in that the at least one detection unit ( 44 ) as an optical detection unit ( 44 ), especially as a digital camera ( 46 ), is trained. Soil cultivation device according to claim 21 or 22, characterized in that the soil cultivation device ( 10 ) a pattern recognition element ( 54 ) for extracting features of the images and for constructing comparison patterns, and a comparison element ( 56 ) for comparing the comparison patterns with one or more in the memory element ( 58 ) stored reference patterns. Soil cultivation device according to claim 23, characterized in that the soil cultivation device ( 10 ) a localization member ( 78 ) for determining the position of the soil tillage implement ( 10 ) relative to the bottom surface segments having the basic pattern on the basis of comparative data of the comparison element, which are obtained by comparing two or more successive images of the floor surface ( 12 ) are recoverable. Soil cultivation device according to one of claims 21 to 24, characterized in that the soil cultivation device in each case a drive wheel ( 15 . 16 ) associated Radencoder ( 82 . 83 ) has for determining the number of revolutions of the respective drive wheel ( 15 . 16 ). Soil cultivation device according to one of claims 21 to 25, characterized in that the soil cultivation device ( 10 ) a computing element ( 80 ) for determining the size of a bottom surface segment comprising the basic pattern and / or for determining the size of the soil cultivation device ( 10 ) traveled distance. Soil cultivation device according to one of claims 21 to 26, characterized in that the soil cultivation device ( 10 ) one or more sensors ( 96 . 98 ) for detecting obstacles on the ground surface ( 12 ) and / or to detect limitations ( 99 . 100 ) of the floor surface ( 12 ). Soil cultivation device according to one of claims 21 to 27, characterized in that the soil cultivation device ( 10 ) a display unit ( 110 ) for displaying an image ( 118 ) the map ( 72 ). Soil cultivation device according to one of claims 21 to 28, characterized in that the soil cultivation device ( 10 ) is self-propelled and self-steering and a drive unit ( 25 ) and a control unit ( 24 ) for controlling the movement of the cultivator ( 10 ) having. Soil cultivation device according to one of claims 21 to 29, characterized in that the soil cultivation device ( 10 ) a floor cleaning device ( 10 ).

Images