WO2019104732A1

WO2019104732A1 - Vision cleaning robot and obstacle detection method

Info

Publication number: WO2019104732A1
Application number: PCT/CN2017/114323
Authority: WO
Inventors: 王声平; 张立新
Original assignee: 深圳市沃特沃德股份有限公司
Priority date: 2017-12-01
Filing date: 2017-12-01
Publication date: 2019-06-06

Abstract

An obstacle detection method comprises: constructing a spatial three-dimensional coordinate system, in which the feature points are located, so as to form a spatial feature point set; extracting, according to the spatial feature point set, necessary feature points in accordance with specified conditions, so as to form a spatial obstacle point set; and projecting the spatial obstacle point set onto a two-dimensional plane, so as to obtain the coordinate points of the spatial obstacle point set on the two-dimensional plane, and detect the coordinate position of an obstacle in accordance to said coordinate points, thereby assisting in prediction of the obstacle.

Description

\¥0 2019/104732 卩(:17 \2017/114323

1 Vision sweeping robot and obstacle detection method

[0001] The present invention relates to the field of visual sweeping robots, and more particularly to a visual sweeping robot and an obstacle detecting method.

Background technique

[0002] When the visual sweeping robot cleans the home area, there are unavoidable obstacles such as sofas, tables, and cabinets in the home area. When cleaning, it is necessary to avoid these obstacles. Therefore, the visual sweeping robot needs to detect obstacles in the visual range in advance.

[0003] Currently, methods for detecting an obstacle include a method using infrared reflection, a method using ultrasonic detection, a method of physical collision, and the like. The current methods all have corresponding defects. For example, the method of infrared reflection is not sensitive to black obstacles; in the ultrasonic detection method, it is difficult to install ultrasonic waves, and the obstacle distance cannot be measured due to insufficient reflection of the surface of the narrow-shaped obstacle; The method of physical collision has a large jitter of the visual sweeping robot during the collision, which affects the stability of the visual sweeping robot.

technical problem

[0004] A primary object of the present invention is to provide a visual cleaning robot and an obstacle detecting method for detecting a coordinate position of an obstacle to assist in predicting an obstacle.

Problem solution

Technical solution

[0005] The present invention provides an obstacle detection method, which includes the following steps:

[0006] acquiring image information;

[0007] extracting feature points in the image information, the feature points including necessary feature points;

[0008] constructing a spatial three-dimensional coordinate of the feature point to form a spatial feature point set;

[0009] extracting necessary feature points according to the specified condition to form a set of spatial obstacle points according to the set of spatial feature points; [0010] projecting the set of spatial obstacle points onto a two-dimensional plane to obtain the spatial barrier The coordinate points of the object points on a two-dimensional plane.

[0011] The present invention also provides a cleaning robot, comprising: \¥0 2019/104732 卩(:17 \2017/114323

2

[0012] Meter set single 兀, used for meter image information,

[0013] an extracting unit, configured to extract a feature point in the image information, where the feature point includes a necessary feature point; [0014] a building unit, configured to construct a spatial three-dimensional coordinate of the feature point to form a spatial feature point [0015] a feature unit, configured to extract necessary feature points according to the specified condition according to the set of spatial feature points to form a set of spatial obstacle points;

[0016] a projection unit, configured to project the set of spatial obstacle points onto a two-dimensional plane to obtain coordinate points of the spatial obstacle point set on a two-dimensional plane.

Advantageous effects of the invention

Beneficial effect

[0017] The visual cleaning robot and the obstacle detecting method provided by the present invention have the following beneficial effects:

[0018] The visual cleaning robot and the obstacle detecting method provided by the present invention, by acquiring image information; extracting feature points in the image information, the feature points include necessary feature points; constructing a spatial three-dimensional space in which the feature points are located Coordinates to form a set of spatial feature points; according to the set of spatial feature points, extract necessary feature points according to specified conditions to form a set of spatial obstacle points; project the set of spatial obstacle points onto a two-dimensional plane to obtain the The coordinate points of the space obstacle points on the two-dimensional plane; thereby detecting the coordinate position of the obstacle to assist in predicting the obstacle.

Brief description of the drawing

DRAWINGS

1 is a schematic diagram showing the steps of an obstacle detecting method according to an embodiment of the present invention;

2 is a schematic diagram showing specific steps of step 32 in an embodiment of the present invention;

3 is a schematic diagram showing steps of an obstacle detecting method according to another embodiment of the present invention;

4 is a schematic diagram showing the steps of an obstacle detecting method according to another embodiment of the present invention;

5 is a schematic diagram showing the specific steps of step 39 in an embodiment of the present invention;

6 is a schematic diagram showing the specific steps of step 39 in another embodiment of the present invention;

7 is a schematic diagram showing the specific steps of step 34 in another embodiment of the present invention;

8 is a block diagram showing the structure of a visual sweeping robot according to an embodiment of the present invention;

9 is a block diagram showing the structure of an extracting unit in an embodiment of the present invention;

10 is a block diagram showing the structure of a visual sweeping robot according to another embodiment of the present invention; \¥0 2019/104732 卩(:17 \2017/114323

3

11 is a block diagram showing the structure of a processing unit in another embodiment of the present invention;

12 is a block diagram showing the structure of a processing unit in still another embodiment of the present invention;

13 is a block diagram showing the structure of a processing unit in still another embodiment of the present invention;

14 is a block diagram showing the structure of a visual sweeping robot according to still another embodiment of the present invention;

[0033] FIG. 15 is a block diagram showing the structure of a feature unit in an embodiment of the present invention.

[0034] The implementation, functional features, and advantages of the present invention will be further described with reference to the accompanying drawings.

BEST MODE FOR CARRYING OUT THE INVENTION

[0035] It is to be understood that the specific embodiments described herein are merely illustrative of the invention.

[0036] The singular forms "a," ""," It is to be understood that the phrase "comprises" or "an" Other characteristics, integers, steps, operations, components, units, modules, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element. Further, "connected" or "coupled" as used herein may include either a wireless connection or a wireless coupling. The term "and/or" used herein includes all or any of the elements and all combinations of one or more of the associated listed.

1 is a schematic diagram of steps of an obstacle detection method according to an embodiment of the present invention.

[0038] An embodiment of the present invention provides an obstacle detection method for a visual sweeping robot, the method comprising the following steps:

[0039] Step 31: acquiring image information;

[0040] Step 32: Extract feature points in the image information, where the feature points include necessary feature points;

[0041] Step 33: Construct a spatial three-dimensional coordinate of the feature point to form a spatial feature point set;

[0042] Step 34: Extract, according to the set of spatial feature points, the necessary feature points according to the specified conditions to form a set of spatial obstacle points;

[0043] Step 35, projecting the set of spatial obstacle points onto a two-dimensional plane to obtain the spatial obstacle point \¥0 2019/104732 卩(:17 \2017/114323

4 sets the coordinate points on the 2D plane.

[0044] In the embodiment, during the cleaning process of the visual sweeping robot, a small range of movement is performed, and the image information in the visual range is continuously collected by the built-in camera, and the image information is an image including a ceiling, a ground, and a wall surface. Information; feature points are pixels, which can be pixels with higher brightness or darker colors; or pixels with larger changes.

[0045] Specifically, referring to FIG. 2, the step of extracting feature points in the image information includes:

[0046] 801: Arranging a plurality of collected pictures or a plurality of frames of video in chronological order;

[0047] 802: analyzing whether there is a pixel change in the same direction or the same route in the image information in the plurality of pictures or the multi-frame video arranged in chronological order;

[0048] 803: analyzing whether the change of the pixel point is greater than a preset value;

[0049] In this step, it may be analyzed whether the size of the same direction or the same route pixel point change in the image information in the adjacent two pictures or two frames of video is greater than a preset value, and the preset value may be 3111111, for example, comparing the latter Whether the diameter of the pixel in the picture differs from the diameter of the pixel in the previous picture is greater than 3111111, and if it is greater, step 804 is performed.

[0050] 804: If it is greater, the pixel is extracted.

[0051] after extracting the feature points of the image information in the visual range, reconstructing the spatial point information of the cleaning environment by using a method of visual SLAM (real-time positioning and map construction to realize autonomous positioning and navigation of the robot) to construct the feature points The three-dimensional coordinates of the space in which it is located. The above feature points necessarily include unnecessary feature points of the ceiling position, and it is impossible to become an obstacle of the visual sweeping robot because the ceiling or the like. Therefore, the position of the ceiling is analyzed by the method of plane fitting, and the non-essential feature points of the ceiling and its vicinity (such as the distance of 1111) are all eliminated, and the necessary feature points are obtained, and the necessary feature points are formed in the three-dimensional coordinates of the space. A set of space obstacles.

[0052] Finally, the necessary feature points on the three-dimensional coordinates of the space are projected on the two-dimensional plane, and the coordinate points of the spatial obstacle point set on the two-dimensional plane are obtained. The obstacle detection method in this embodiment has the advantages of high applicability, high detection accuracy, low cost, and low computational complexity.

[0053] Referring to FIG. 3, in an embodiment, the step of projecting the set of spatial obstacle points onto a two-dimensional plane to obtain coordinate points of the spatial obstacle point set on a two-dimensional plane is performed. After that, including:

[0054] Step 36: Mark the coordinate point according to the coordinate point of the spatial obstacle point set on the two-dimensional plane \¥0 2019/104732 卩(:17 \2017/114323

5 The position corresponding to the 2D map of the cleaning area.

[0055] In this embodiment, the coordinate origin of the two-dimensional plane is the same as the origin of the two-dimensional map, and the coordinate point is conveniently marked in the corresponding position of the two-dimensional map. The solution in this embodiment is convenient for the sweeping robot to press the cleaning area. The coordinate point can be controlled to bypass when walking on the map. The map of the cleaning area may be pre-established and stored in the visual sweeping robot, or may be a map created by the visual sweeping robot when cleaning.

[0056] Referring to FIG. 4, in an embodiment, the step of projecting the set of spatial obstacle points onto a two-dimensional plane to obtain coordinate points of the spatial obstacle point set on a two-dimensional plane is performed. After that, include:

[0057] Step 37: Acquire current positioning coordinates;

[0058] Step 38: analyzing a positional relationship between the current positioning coordinate and the coordinate point;

[0059] Step 39: Select a corresponding preset manner to perform processing according to the location relationship.

[0060] After acquiring the coordinate points of the spatial obstacle point set on the two-dimensional plane, the visual sweeping robot acquires the current positioning from time to time, and marks the current positioning on the two-dimensional plane to obtain the current positioning on the two-dimensional plane. Coordinates, the visual sweeping robot can perform processing according to the positional relationship between its current positioning coordinates and the coordinate points of the space obstacle point set, for example, avoid obstacles, and explore all obstacles in the cleaning area.

[0061] Specifically, referring to FIG. 5, in an embodiment, the step 39 of selecting a corresponding preset manner according to the location relationship, includes:

[0062] Step 391, determining whether the location relationship is less than a preset value;

[0063] Step 392, if yes, controlling the deceleration operation.

[0064] The positional relationship is the distance between the current positioning coordinate and the coordinate point of the spatial obstacle point set on the two-dimensional plane, and the preset value is 0.5 m.

[0065] In another embodiment, after the step 392 of controlling the deceleration operation, the method includes:

[0066] Step 393, analyzing feature attributes of the set of spatial obstacle points;

[0067] The feature attribute is the size of the length, width, and height of the set of spatial obstacle points.

[0068] Step 394: Analyze whether the feature attribute is greater than a preset attribute; and the preset attribute is a preset size of a length, a width, and a height.

[0069] Step 395, if it is greater, controlling to avoid the space obstacle point set.

[0070] Preferably, in this embodiment, the space obstacle point set may also be marked by the feature attribute in the cleaning area. \¥0 2019/104732 卩(:17 \2017/114323

6 on the map.

[0071] Referring to FIG. 6 , in another embodiment, the plurality of spatial obstacle points are set, and correspondingly, the position relationship is multiple, and the position relationship is a position distance, according to the Step 39 of processing the preset relationship corresponding to the location relationship, including:

[0072] Step 3901: Select a shortest position distance from a plurality of the location distances;

[0073] Step 3902: Control a spatial obstacle point set motion corresponding to the shortest position distance.

[0074] After step 3902, the method includes:

[0075] Step 3903, analyzing whether the shortest position distance is less than a predetermined value;

[0076] Step 3904, if less than, starting the ultrasonic sensor to emit an ultrasonic signal;

[0077] Step 3905, receiving a feedback signal that is fed back when the ultrasonic signal encounters an obstacle;

[0078] Step 3906, analyzing location information and feature attributes of the obstacle according to the feedback signal;

[0079] Step 3907, analyzing whether the feature attribute is greater than a preset attribute;

[0080] Step 3908, if greater than, clear the mark corresponding to the coordinate point of the space obstacle point set on the two-dimensional map, and mark the position information of the obstacle in the corresponding position of the two-dimensional map of the cleaning area;

[0081] Step 3910: Store a two-dimensional map marked with the location information of the obstacle, so as to avoid the obstacle when designing the cleaning route.

[0082] Further, after the step 36 of marking the coordinate point on the position corresponding to the two-dimensional map of the cleaning area according to the coordinate point of the spatial obstacle point set on the two-dimensional plane, the method includes:

[0083] According to the mark of the coordinate point on the two-dimensional map, the cleaning area is divided into regions.

[0084] After detecting all the sets of spatial obstacles in the cleaning area, the cleaning area may be divided according to the position of the coordinate points of all the spatial obstacle points in the cleaning area. Preferably, the sub-area after the area division may also be subjected to map coverage.

[0085] Referring to FIG. 7, the feature points further include non-essential feature points, and the step 34 of extracting the necessary feature points according to the specified set of spatial feature points to form a set of spatial obstacle points according to the set of spatial feature points includes:

[0086] Step 341, analyzing a positional relationship of the feature points in the set of spatial feature points;

[0087] Step 342, distinguishing the non-essential feature points and the necessary feature points according to the positional relationship;

[0088] Step 343: cull the non-essential feature points according to the specified condition;

[0089] Step 344, forming the necessary feature points into a set of spatial obstacle points. \¥0 2019/104732 卩(:17 \2017/114323

7

[0090] The condition specified in this embodiment is to retain the necessary feature points, since the ceiling or the like cannot be an obstacle to the visual sweeping robot. Therefore, the visual sweeping robot of the present embodiment analyzes the position of the ceiling by using a plane fitting method, and removes all unnecessary feature points of the ceiling and its preset range (such as the distance of 1111) to obtain necessary feature points. The necessary feature points form a set of spatial obstacle points in the three-dimensional coordinates of the space.

[0091] In summary, the method for detecting an obstacle provided in the embodiment of the present invention, by acquiring image information; extracting feature points in the image information, the feature points include non-essential feature points and necessary feature points; Constructing a spatial three-dimensional coordinate of the feature point to form a spatial feature point set; according to the spatial feature point set, removing the necessary feature point according to a specified condition, and obtaining a necessary feature point to form a spatial obstacle point set; The obstacle point set is projected onto the two-dimensional plane to obtain coordinate points of the spatial obstacle point set on the two-dimensional plane; thereby detecting the coordinate position of the obstacle to assist in predicting the obstacle. It can also detect unknown environments and divide the cleaning area.

[0092] Referring to FIG. 8, an embodiment of the present invention further provides a visual cleaning robot, including:

[0093] The collecting unit 10 is configured to collect image information in a visual range;

[0094] The extracting unit 20 is configured to extract feature points in the image information, where the feature points include necessary feature points

[0095] a building unit 30, configured to construct a spatial three-dimensional coordinate of the feature point to form a spatial feature point set;

[0096] a feature unit 40, configured to extract, according to the set of spatial feature points, necessary feature points according to a specified condition to form a set of spatial obstacle points;

[0097] The projection unit 50 is configured to project the set of spatial obstacle points onto a two-dimensional plane to obtain coordinate points of the spatial obstacle point set on a two-dimensional plane.

[0098] In the embodiment, during the cleaning process of the visual sweeping robot, a small range of movement is performed, and the collecting unit 10 continuously collects image information in the visual range through the built-in camera, and the image information includes the ceiling, the ground, and the wall surface. Image information. The feature points are pixels, which can be pixels with higher brightness or darker colors.

Referring to FIG. 9, the extracting unit 20 of the present invention includes:

[0100] Arrangement unit 201: used to sequentially arrange multiple captured pictures or multiple frames of video in chronological order \¥0 2019/104732 卩(:17 \2017/114323

8

[0101] The first analyzing unit 202 is configured to analyze whether there is a pixel point change in the same direction or the same route in the image information in the plurality of pictures or the multi-frame video arranged in chronological order;

[0102] The second analyzing unit 203 is configured to analyze whether the change of the pixel point is greater than a preset value;

[0103] The second analyzing unit 203 may analyze whether the size of the same direction or the same route pixel point change in the image information in the adjacent two pictures or two frames of video is greater than a preset value, and the preset value may be 3 mm, for example, The second analyzing unit 203 compares whether the diameter of the pixel in the next picture differs from the diameter of the pixel in the previous picture by more than 3 mm.

[0104] The extracting unit 204: extract the pixel point if the change of the pixel point is greater than a preset value.

[0105] After extracting the feature points of the image information in the visual range, the constructing unit 30 reconstructs the spatial point information of the cleaning environment by using a method of visual SLAM (real-time positioning and map construction to realize autonomous positioning and navigation of the robot) to construct the The three-dimensional coordinates of the space where the feature points are located. The above feature points necessarily include non-essential feature points of the ceiling plate, and it is impossible to become an obstacle of the visual sweeping robot because the ceiling or the like. Therefore, the feature unit 40 uses the method of plane fitting to analyze and estimate the position of the ceiling, and eliminates all unnecessary feature points such as the ceiling and its vicinity (such as within the range of lm distance) to obtain necessary feature points, and the necessary feature points are A set of spatial obstacle points is formed in the three-dimensional coordinates of the space.

[0106] Finally, the projection unit 50 projects the three-dimensional coordinates of the necessary feature points in the above-mentioned spatial obstacle points on the two-dimensional plane, and obtains the coordinate points of the obstacle on the two-dimensional plane. The visual sweeping robot in this embodiment has the advantages of high applicability, high detection accuracy, low cost, and low computational complexity.

[0107] In an embodiment, the visual cleaning robot further includes:

[0108] The marking unit 51 is configured to mark the coordinate point at a position corresponding to the two-dimensional map of the cleaning area according to the coordinate point of the spatial obstacle point set on the two-dimensional plane.

[0109] In this embodiment, the coordinate origin of the two-dimensional plane is the same as the origin of the two-dimensional map, and the convenient marking unit 51 quickly marks the coordinate point at a position corresponding to the two-dimensional map, and the map of the cleaning area may be pre-established and stored. In the visual sweeping robot, it is also a map created when the visual sweeping robot is cleaned.

[0110] Referring to FIG. 10, in an embodiment, the visual cleaning robot further includes:

[0111] an obtaining unit 60, configured to acquire current positioning coordinates;

[0112] The analyzing unit 70 is configured to analyze a positional relationship between the current positioning coordinate and the coordinate point;

[0113] The processing unit 80 is configured to perform processing according to the location relationship to select a corresponding preset manner. \¥0 2019/104732 卩(:17 \2017/114323

9

[0114] The acquiring unit 60 acquires the current positioning from time to time, and marks the current positioning on the two-dimensional plane to obtain the current positioning coordinates on the two-dimensional plane, and the analyzing unit 70 analyzes the current positioning coordinate and the coordinate on the two-dimensional plane. The positional relationship of the points, the processing unit 80 can perform corresponding processing according to the positional relationship between the current positioning coordinates of the visual cleaning robot itself and the coordinate points of the spatial obstacle point set, for example, avoid obstacles, and explore all obstacles in the cleaning area. Things and so on.

[0115] Specifically, referring to FIG. 11, in an embodiment, the processing unit 80 includes:

[0116] a determining subunit 801, configured to determine whether the location relationship is less than a preset value;

[0117] The deceleration subunit 802 is configured to control the deceleration operation when the positional relationship is less than a preset value.

[0118] The positional relationship is the distance between the current positioning coordinate and the coordinate point of the spatial obstacle point set on the two-dimensional plane, and the preset value is 0.5 m.

[0119] In another embodiment, referring to FIG. 12, the processing unit 80 further includes:

[0120] a first analysis subunit 803, configured to analyze feature attributes of the set of spatial obstacle points;

[0121] The feature attribute is the size of the length, width, and height of the set of spatial obstacle points.

[0122] The second analysis sub-unit 804 is configured to analyze whether the feature attribute is greater than a preset attribute; and the preset attribute is a preset size of a length, a width, and a height.

[0123] The avoidance sub-unit 805 is configured to control to avoid the spatial obstacle object point set if the feature attribute is greater than a preset attribute.

[0124] Preferably, in this embodiment, the feature attribute of the spatial obstacle point set may also be marked on the map of the cleaning area.

[0125] Referring to FIG. 13 , in another embodiment, the plurality of spatial obstacle points are set, and correspondingly, the position relationship is multiple, and the position relationship is a position distance, and the processing unit 80 Includes:

[0126] a selection subunit 810, configured to select a shortest position distance from a plurality of the location distances;

[0127] The control subunit 820 is configured to control the space obstacle point set motion corresponding to the shortest position distance.

[0128] The processing unit 80 further includes:

[0129] The distance analyzing unit 821 is configured to analyze whether the shortest position distance is less than a predetermined value;

[0130] The starting unit 822 is configured to: when the shortest position distance is less than a predetermined value, start the ultrasonic sensor to emit an ultrasonic signal; \¥0 2019/104732 卩(:17 \2017/114323

10

[0131] The signal receiving unit 823 is configured to receive a feedback signal that is fed back when the ultrasonic signal encounters an obstacle;

[0132] The information analysis unit 824 is configured to analyze location information and feature attributes of the obstacle according to the feedback signal. [0133] an attribute analysis unit 825, configured to analyze whether the feature attribute is greater than a preset attribute;

[0134] The obstacle marking unit 826 is configured to: if the feature attribute is greater than the preset attribute, clear the mark corresponding to the coordinate point of the space obstacle point set on the two-dimensional map, and mark the position information of the obstacle in the two-dimensional map of the cleaning area. Corresponding location

[0135] The location storage unit 827 is configured to store a two-dimensional map of the location information marked with the obstacle so as to avoid the obstacle when planning the cleaning route.

[0136] Further, referring to FIG. 14, there are a plurality of coordinate points, and the visual cleaning robot further includes:

[0137] a dividing unit 61, configured to divide the cleaning area according to the marking of the plurality of coordinate points on the map

[0138] After all the spatial obstacle points in the cleaning area are detected, the cleaning area may be divided according to the position of the coordinate point corresponding to all the spatial obstacle points in the cleaning area. Preferably, the sub-area after the area division can also be covered by the map.

[0139] Further, the feature points further include non-essential feature points.

[0140] Referring to FIG. 15, the feature unit 40 includes:

[0141] a third analysis subunit 401, configured to analyze a positional relationship of the feature points in the set of spatial feature points

[0142] a region numerator unit 402, configured to distinguish the non-essential feature point and the necessary feature point according to the positional relationship;

[0143] a culling sub-unit 403, configured to cull the non-essential feature points according to the specified condition;

[0144] Forming a subunit 404, configured to form the necessary feature points into a set of spatial obstacle points.

[0145] The condition specified in the present embodiment is to retain the necessary feature points, since the ceiling or the like is unlikely to be an obstacle to the visual sweeping robot. Therefore, the region molecular unit 402 analyzes the non-denot feature points such as the position of the ceiling from the feature points by using a plane fitting method, and the culling sub-unit 403 sets the non-essential feature points of the ceiling and its preset range (eg, within the distance of 1111). All are eliminated, and the necessary feature points are obtained. The forming sub-unit 404 forms the set of spatial obstacle points in the three-dimensional coordinates of the space.

[0146] In summary, the method for detecting a visual sweeping robot and an obstacle detected by the embodiment of the present invention is \¥0 2019/104732 卩(:17 \2017/114323

11 acquiring image information; extracting feature points in the image information, the feature points include necessary feature points; constructing spatial three-dimensional coordinates of the feature points to form a spatial feature point set; according to the spatial feature point set, according to Specifying conditions to remove the necessary feature points, obtaining necessary feature points to form a spatial obstacle point set; projecting the spatial obstacle point set onto the two-dimensional plane to obtain the spatial obstacle point set on the two-dimensional plane Coordinate point; thereby detecting the coordinate position of the obstacle to assist in predicting the obstacle; and detecting the unknown environment and dividing the cleaning area.

The above is only a preferred embodiment of the present invention, and is not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the invention and the drawings are used directly or indirectly. Other related technical fields are equally included in the scope of patent protection of the present invention.

Claims

\¥0 2019/104732 卩(:17 \2017/114323 12 Claims

[Claim 1] An obstacle detecting method, comprising the steps of:

Collecting image information;

Extracting feature points in the image information, the feature points include necessary feature points; constructing spatial three-dimensional coordinates of the feature points to form a spatial feature point set; and extracting necessary features according to the specified conditions according to the spatial feature point set Point to form a set of spatial obstacle points;

The spatial obstacle point set is projected onto a two-dimensional plane to obtain coordinate points of the spatial obstacle point set on a two-dimensional plane.

[Claim 2] The obstacle detecting method according to claim 1, wherein the spatial obstacle point set is projected onto a two-dimensional plane to obtain the spatial obstacle point set in two dimensions After the steps of the coordinate points on the plane, include:

The coordinate points are marked at positions corresponding to the two-dimensional map of the cleaning area according to the coordinate points of the spatial obstacle point set on the two-dimensional plane.

[Claim 3] The obstacle detecting method according to claim 1, wherein the spatial obstacle point set is projected onto a two-dimensional plane to obtain the spatial obstacle point set in two dimensions After the steps of the coordinate points on the plane, include:

Get the current positioning coordinates;

Analyzing a positional relationship between the current positioning coordinate and the coordinate point;

The corresponding preset manner is selected according to the positional relationship for processing.

[Claim 4] The obstacle detection method according to claim 3, wherein the preset manner includes a deceleration operation, and the step of selecting a corresponding preset manner to perform processing according to the positional relationship includes:

Determining whether the positional relationship is less than a preset value;

If yes, control the deceleration operation.

[Claim 5] The obstacle detecting method according to claim 4, wherein the step of controlling the deceleration operation comprises:

Analyzing characteristic attributes of the set of spatial obstacle points; \¥0 2019/104732 卩(:17 \2017/114323

13 analyzing whether the feature attribute is greater than a preset attribute;

If it is greater, the control avoids the set of space obstacle points.

[Claim 6] The obstacle detecting method according to claim 3, wherein the plurality of spatial obstacle points are plural, and correspondingly, the positional relationship is plural, and the positional relationship is a position. The step of selecting the corresponding preset manner for processing according to the location relationship includes:

Selecting the shortest position distance from a plurality of the position distances;

Controlling the movement of the point set of the space obstacle corresponding to the shortest position distance.

[Claim 7] The obstacle detecting method according to claim 2, wherein the feature attribute is a length, a width, a height, and a preset attribute is a preset length, width, and height.

[Claim 8] The obstacle detecting method of the visual cleaning robot according to claim 1, wherein the feature point further includes a non-essential feature point, and the extracting is performed according to the specified condition according to the spatial feature point set The step of forming a feature point to form a set of spatial obstacle points includes: analyzing a positional relationship of the feature points in the set of spatial feature points;

Distinguishing the non-essential feature points and the necessary feature points according to the positional relationship; culling the non-essential feature points according to the specified condition;

The necessary feature points are formed into a set of spatial obstacle points.

[Claim 9] A cleaning robot, comprising:

Acquisition unit for collecting image information,

An extracting unit, configured to extract feature points in the image information, where the feature points include necessary feature points;

a building unit, configured to construct a spatial three-dimensional coordinate of the feature point to form a spatial feature point set;

a feature unit, configured to extract necessary feature points according to the specified condition according to the set of spatial feature points to form a set of spatial obstacle points;

a projection unit, configured to project the set of spatial obstacle points onto a two-dimensional plane to obtain coordinate points of the set of spatial obstacle points on a two-dimensional plane.

The visual cleaning robot according to claim 9, further comprising: \¥0 2019/104732 卩(:17 \2017/114323

The marking unit marks the coordinate point at a position corresponding to the two-dimensional map of the cleaning area according to the coordinate point of the spatial obstacle point set on the two-dimensional plane.

The visual cleaning robot according to claim 9, further comprising:

An obtaining unit, configured to acquire a current positioning coordinate;

The analyzing unit is configured to analyze a positional relationship between the current positioning coordinate and the coordinate point; and the processing unit is configured to select a corresponding preset manner to perform processing according to the positional relationship.

[Claim 12] The visual cleaning robot according to claim 11, wherein the preset mode comprises a deceleration operation, and the processing unit comprises:

a determining subunit, configured to determine whether the positional relationship is less than a preset value;

The deceleration subunit, when the positional relationship is less than a preset value, controls the deceleration operation.

[Claim 13] The visual cleaning robot according to claim 12, wherein the processing unit further comprises:

a first analysis subunit, configured to analyze feature attributes of the set of spatial obstacle points; a second analysis subunit, configured to analyze whether the feature attribute is greater than a preset attribute; and a avoidance subunit, configured to use the feature attribute If it is greater than the preset attribute, the control avoids the space obstacle point set.

The visual cleaning robot according to claim 11, wherein the plurality of spatial obstacle points are plural, and correspondingly, the positional relationship is plural, and the positional relationship is a positional distance. The processing unit includes:

Selecting a subunit for selecting a shortest position distance from a plurality of the position distances; and a control subunit for controlling a movement of the point crap set corresponding to the shortest position distance.

[Claim 15] The visual cleaning robot according to claim 10, wherein the feature attribute is a length, a width, and a height, and the preset attribute is a preset length, width, and height.

[Claim 16] The visual cleaning robot according to claim 9, wherein the feature point further includes a non-essential feature point, and the feature unit includes:

a third analysis subunit, configured to analyze a positional relationship of the feature points in the set of spatial feature points; \¥0 2019/104732 卩(:17 \2017/114323

a 15-segment molecular unit for distinguishing the non-essential feature points and the necessary feature points according to the positional relationship;

The culling subunit is configured to cull the non-essential feature points according to the specified condition; and form a subunit for forming the necessary feature points into a set of spatial obstacle points.