WO2021139683A1

WO2021139683A1 - Self-moving device

Info

Publication number: WO2021139683A1
Application number: PCT/CN2021/070477
Authority: WO
Inventors: 朱松; 何明明
Original assignee: 苏州宝时得电动工具有限公司
Priority date: 2020-01-06
Filing date: 2021-01-06
Publication date: 2021-07-15
Also published as: WO2021139414A1; CN114868095A

Abstract

A self-moving device (100), comprising: a housing (110), a moving module (130), a working module (102), an image collecting apparatus (140), and a control module (101) which is used to autonomously control the moving module (130) to drive the housing (110) to move, and which autonomously controls the working module (102) to execute a preset working task. The self-moving device (100) comprises an edge mode, and in the edge mode, the control module (101) is configured to: according to environmental images, collected by the image collecting apparatus (140), of where same is located, determine whether the ratio of non-working region to working region in the environmental images exceeds a threshold, and if exceeding, then perform boundary fitting on the working region in the environmental images to form a fit boundary line (31); generate a target point (D) according to the fit boundary line (31); control the self-moving device (100) to move to the target point (D); and control the self-moving device (100) to move along the boundary of the working region.

Description

Self-mobile

This application requires that the application date is January 6, 2020, the application number is 202010011200.2, the application date is June 19, 2020, the application number is 202010565814.5, and the application date is July 6, 2020, the application number is 202010642119.4 China The priority of the patent application, the entire content of which is incorporated in this application by reference.

Technical field

The invention relates to a self-moving device.

Background technique

With the continuous advancement of computer technology and artificial intelligence technology, automatic lawn mowers similar to intelligent robots have begun to slowly enter people's lives. For example, an automatic lawn mower can automatically mow and charge the user's lawn without user intervention. After this automatic working system is set up once, there is no need to invest in management, freeing users from boring, time-consuming and laborious housework such as cleaning and lawn maintenance. At present, the automatic lawn mower moves randomly within the working area defined by the boundary line, but it is cumbersome to arrange the boundary line. Therefore, it is necessary to design a new automatic lawn mower to solve the above-mentioned problems.

Summary of the invention

In order to overcome the above drawbacks, the present invention adopts the following technical solutions:

A self-moving device, including:

case;

The moving module is located below the housing and is used to drive the housing to move;

The working module is arranged on the housing to perform preset working tasks;

An image acquisition device for acquiring an image of the environment in which the self-mobile device is located;

The control module is configured to autonomously control the movement module to drive the housing to move, and autonomously control the working module to perform preset working tasks;

The self-moving device includes an edge mode, and in the edge mode, the control module is configured to:

According to the environment image collected by the image acquisition device, determine whether the proportion of the non-working area to the working area in the environment image exceeds the threshold; if it exceeds, then:

Fitting the boundary of the working area in the environmental image into a fitting boundary line;

Generating a target point according to the fitted boundary line;

Controlling the self-mobile device to move to the target point;

Controlling the self-moving device to move along the boundary of the work area.

Further, a parallel line separated by a preset distance from the fitting boundary line is defined as a target line, and the target point is a point on the target line.

Further, the generating a target point according to the fitted boundary line includes: selecting a point in the fitted boundary line as a target base point, and a line connecting the target point and the target base point is perpendicular to the target base point. Fit the boundary line, and the distance between the target point and the target base point is a preset distance.

Further, defining a point on the self-mobile device as a fitting point of the self-mobile device, and controlling the self-mobile device to move to the target point includes: controlling the self-mobile device to move so that the Move from the fitting point of the mobile device to the target point.

Further, the preset distance is equal to the distance from the fitting point to the side of the mobile device, or equal to the distance from the fitting point to the side of the mobile device plus a safe distance.

Further, the controlling the self-mobile device to move to the target point further includes:

Rotate the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line.

Further, controlling the movement of the self-mobile device so that the fitting point of the self-mobile device moves to the target point includes: controlling the self-mobile device to move between the fitting point and the target point The line is moved to move the fitting point to the target point.

Further, the rotating the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line includes: according to the connection between the fitting point and the target point and Rotating the self-moving device at the included angle of the fitted straight line so that the traveling direction of the self-moving device is the same as the extending direction of the fitted boundary line.

Further, the image acquisition device includes a side image acquisition device located on one side of the self-moving device, and the self-moving device is rotated so that the driving direction of the self-moving device is consistent with the fitting boundary line. The extension direction of is the same, including: rotating the self-moving device to the side where the side image acquisition device is located close to the boundary of the working area.

Further, the controlling the self-moving device to move along the boundary of the working area includes:

Controlling the self-mobile device to move forward and synchronously collecting current environment images;

Fitting the boundary of the working area in the current environment image to the current fitting boundary line;

Controlling the self-moving device to move along the current fitting boundary line.

Further, the fitting the boundary of the working area in the current environment image to the current fitting boundary line includes:

Divide the current environment image into N sub-images along the distance from the mobile device, where N≥2;

Fitting the boundary of the working area in each of the sub-images into a straight boundary line;

Control the self-moving device to move along the linear boundary line synthesized by the sub-image closest to the self-mobile device, and the linear boundary line synthesized according to the remaining sub-images is the boundary line of the self-mobile device Make predictions for subsequent sports.

Further, N=2.

Further, the control module is configured to:

Before judging whether the proportion of the non-working area to the working area in the environmental image exceeds a threshold value according to the environment image collected by the image acquisition device, the method further includes: according to the environment image collected by the image acquisition device, Analyze whether there is a boundary of the working area in the environment image, and when the boundary of the working area does not exist in the environment image, control the moving module to move according to a preset edge finding logic to find the boundary of the working area; When the working area boundary exists in the environment image, it is determined whether the proportion of the non-working area and the working area on both sides of the boundary in the environment image exceeds a threshold according to the environment image collected by the image acquisition device.

Further, the control module is configured to:

If it is determined that the proportion of the non-working area to the working area in the environmental image does not exceed the threshold, the self-moving device is controlled to continue to move forward in the original direction.

Further, in the edge-edge mode, the control module is further configured to determine in real time whether the current working area boundary is lost, and when the current working area boundary is lost, control the moving module to automatically move to find the work Area boundary; when the current work area boundary is not lost, control the self-mobile device to continue to move and work along the current work area boundary.

Further, the real-time determination of whether the boundary of the current working area is lost includes: calculating the ratio of the non-working area to the working area on both sides of the boundary of the current working area, when the difference between the non-working area and the working area is When the ratio is within the preset range, it is determined that the current working area boundary is not lost; when the ratio of the non-working area to the working area is not within the preset range, it is determined that the current working area boundary is lost.

Further, when the boundary of the current working area is lost, controlling the moving module to automatically move to find the boundary of the working area includes: when the boundary of the current working area is lost, controlling the moving module to rotate a certain angle to find the working area. Area boundary; if the rotation does not find the working area boundary, control the moving module to move according to a preset edge finding logic to find the working area boundary.

Further, the self-mobile device further includes an edge detection mode. In the edge detection mode, the control module is configured to: control the image acquisition device to collect the environmental image, and determine the location based on the environmental image. Whether the working area boundary exists in the environment image; when the working area boundary does not exist in the environment image, the movement module is controlled to move and work according to the preset side movement logic; when the environment image exists When the working area boundary is used, the working area boundary in the environment image is fitted into a fitting boundary line, and the parameters of the fitting boundary line are generated, and the self-moving device is controlled according to the parameters Move and work within the boundaries of the work area.

Further, in the edge detection mode, when the working area boundary exists in the environment image, controlling the self-mobile device to move and work within the working area boundary according to the parameter includes: When the image determines that the self-mobile device is close to the boundary of the working area, the self-mobile device is controlled to continue to move a certain distance in the original direction, and then the movement direction of the self-mobile device is adjusted. Further, the self-moving device is an automatic lawn mower that automatically moves and cuts grass on the grass, the working module is a grass mowing module for performing grass cutting tasks, and the working area boundary is the grass boundary.

Further, the image acquisition device includes an edge image acquisition device and an edge detection image acquisition device. In the edge mode, the control module controls the self-moving image according to the environmental image collected by the edge image acquisition device. Movement and cutting of the device; in the edge detection mode, the control module controls the movement and cutting of the self-moving device according to the environmental image collected by the edge detection image acquisition device.

Further, in the left-right direction of the self-moving device, the edge image acquisition device is arranged on the side of the self-moving device close to the boundary of the working area, and the edge detection image acquisition device is arranged on the self-moving device. The side of the mobile device close to the center of the self-mobile device.

Further, in the front-to-back direction, the edge-edge image acquisition device and the edge-detection image acquisition device are both arranged in front of the advancing direction of the self-moving device; in the height direction, the edge-edge image acquisition device and the detection The installation height of the edge image acquisition device does not exceed 20 cm above the ground; in the left and right direction, the distance between the edge image acquisition device and the side of the self-moving device close to the boundary of the working area is S1, and S1 is at 0 Within the range of -5cm, the distance between the edge detection image acquisition device and the central axis of the self-moving device is S2, and S2 is within the range of 0-4cm.

Further, the self-moving device further includes a reference object detection module for detecting a reference object, and the control module controls the self-moving device to automatically be in at least two sub-work areas according to the information detected by the reference object detection module. Move and cut.

Further, when the self-mobile device has finished working in a sub-work area, the control module controls the self-mobile device to control the self-mobile device to search for the reference along the edge according to the environmental image collected by the image acquisition device When the reference object detection module detects the reference object, the control module controls the self-moving device to enter the next working sub-area according to the information detected by the reference object detection module.

The beneficial effect of this solution is that the image acquisition device is used to quickly find the boundary of the working area and move along the boundary of the working area.

The present invention can also adopt the following technical solutions:

A self-moving device that automatically moves and works in the work area, including:

case;

The working module is arranged on the housing to perform preset working tasks;

The self-moving device includes an edge mode, and in the edge mode, the control module is configured to: analyze whether there is an image of the working area in the environment image according to the environment image collected by the image acquisition device. Boundary; when the working area boundary does not exist in the environment image, control the moving module to move according to the preset edge finding logic to find the working area boundary; when the working area boundary exists in the environment image , Divide the environment image into N sub-images, analyze whether the working area boundary exists in each of the sub-images, and respectively fit the working area boundary in each of the sub-images that have the working area boundary Form a straight line, and generate the parameters of the straight line, and control the self-moving device to move and work along the boundary of the working area according to the parameters, where N≥2.

Further, 2≤N≤8.

Further, in the edgewise mode, the control module is further configured to: when the working area boundary exists in the environment image, divide the environment image into a distance from the mobile device into N of the sub-images, and control the movement and operation of the self-mobile device according to the sub-image closest to the self-mobile device, and control the subsequent movement and operation of the self-mobile device according to the remaining sub-images Work to make predictions.

Further, in the edgewise mode, the control module is further configured to: when the working area boundary exists in the environment image, divide the environment image into a distance from the mobile device into Two of the sub-images, and control the movement and work of the self-mobile device according to the sub-image close to the self-mobile device, and control the self-mobile device to move and work according to the sub-image far away from the self-mobile device Make predictions about subsequent moves and work.

Further, the real-time determination of whether the boundary of the current working area is lost includes: counting the proportions of the target and non-standard objects on both sides of the boundary of the current working area. When the ratio of is within the preset range, it is determined that the current working area boundary is not lost; when the ratio of the target object and the non-target object is not within the preset range, it is determined that the current working area boundary is lost.

Further, the self-mobile device further includes an edge detection mode. In the edge detection mode, the control module is configured to: control the image acquisition device to collect the environmental image, and determine the location based on the environmental image. Whether the working area boundary exists in the environment image; when the working area boundary does not exist in the environment image, the movement module is controlled to move and work according to the preset side movement logic; when the environment image exists In the case of the working area boundary, the working area boundary in the environmental image is fitted into a straight line, and the parameters of the straight line are generated, and the self-mobile device is controlled to be at the working area boundary according to the parameters. Move and work inside.

Further, in the edge detection mode, when the working area boundary exists in the environment image, controlling the self-mobile device to move and work within the working area boundary according to the parameter includes: When the image determines that the self-mobile device is close to the boundary of the working area, the self-mobile device is controlled to continue to move a certain distance in the original direction, and then the movement direction of the self-mobile device is adjusted.

Further, the self-moving device is an automatic lawn mower that automatically moves and cuts grass on the grass, the working module is a grass cutting module for performing a mowing task, and the work area boundary is the grass land boundary.

Further, the control module may control the self-mobile device to perform an action corresponding to the current obstacle type according to different types of obstacles in the environment image.

The beneficial effect of this solution is: in the edge mode, the environment image is divided into N sub-images, and the boundary of the working area on each sub-image is respectively fitted into a straight line to generate a parameter representing the straight line, and according to the parameter Controlling the mobile device to automatically move and work along the boundary of the work area can improve the accuracy of the work area boundary obtained from the image, remove the distorted work area boundary, increase the calculation time, and reduce the calculation requirements for the control module. lower the cost.

Description of the drawings

Fig. 1 is a schematic diagram of a self-moving device in an embodiment of the present invention.

Fig. 2 is a block diagram of a self-moving device in an embodiment of the present invention.

Fig. 3 is a block diagram of a self-moving device in an embodiment of the present invention.

Fig. 4 is a schematic diagram of a working state of a self-mobile device in an embodiment of the present invention.

Fig. 5 is a schematic diagram of an edge mode of a self-moving device in an embodiment of the present invention.

Fig. 6 is a schematic diagram of an edge detection mode from a mobile device in an embodiment of the present invention.

Fig. 7 is a schematic diagram of a working state of a self-mobile device in an embodiment of the present invention.

Fig. 8 is a schematic diagram of an environment image taken from a mobile device in Fig. 7.

FIG. 9 is a schematic diagram of the environment image in FIG. 8 being divided into two sub-images.

Fig. 10 is a schematic diagram of an automatic working system in an embodiment of the present invention.

FIG. 11 is a schematic diagram of moving from a mobile device to a target point in an embodiment of the present invention.

Fig. 12 is a schematic diagram of moving from the mobile device to the target point shown in Fig. 11.

FIG. 13 is a schematic diagram from the rotation of the mobile device shown in FIG. 12 until its moving direction is the same as the boundary extension direction.

Detailed ways

In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following further describes the present invention in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, but not used to limit the present invention.

1 to FIG. 13, an embodiment of the present invention provides an automatic working system, which includes a self-mobile device 100 that automatically moves and works in a work area and a charging station for charging the self-mobile device. In this embodiment, the mobile device 100 is an automatic lawn mower, and the charging station is a charging station for charging the automatic lawn mower. In other embodiments, the self-moving device 100 may also be an automatic leaf sweeper, an automatic sprinkler, a multifunction machine, a sweeping robot, and so on.

As shown in FIGS. 1 to 3, the self-mobile device 100 includes a working module 102 arranged in a housing and used to perform preset work tasks, a mobile module 130 located under the housing 110 and used to drive the housing 110 to move, and For the image acquisition device that collects images of the environment in which the mobile device is located, the control module 101 used to control the automatic movement of the mobile module 130 and the automatic operation of the working module 102, and the control module 101 used to supply power to the mobile module, the working module 102 and the control module 101 The energy module 103. The control module 101 is connected to and controls the mobile module 130, the working module 102, the energy module 103, and the image acquisition device 140.

In this embodiment, taking the self-moving device as an automatic lawn mower that automatically moves and cuts grass on the grass as an example, its working area boundary is the grass boundary; its working module 102 is a grass cutting module for performing mowing tasks. , Specifically for cutting parts, such as cutting blades. The mobile module includes auxiliary wheels at the front and driving wheels at the rear. The working module 102 is driven by a cutting motor (not shown). The center of the working module 102 is located on the central axis of the self-moving device 100, located below the housing 110, and located between the auxiliary wheels and the driving wheels.

As shown in FIG. 4, the image capture device 140 is installed on the housing 110 and is used to capture images of the environment where the mobile device is located. In this embodiment, the environment image collected by the image collection device 140 from the mobile device 100 is located, and the environment image refers to the image information of the target area M captured by the image collection device 140. The viewing range of the image capturing device 140 has different viewing angle ranges according to different capturing device types, such as a viewing angle range of 90 degrees to 120 degrees. Of course, in a specific implementation process, a certain angle range within the viewing angle range may be selected as the actual viewing range, for example, a 90-degree range located in the middle within the viewing angle range of 120 degrees may be selected as the actual viewing range. The image acquisition device 140 collects ground visual information obliquely downwards, and the control module 101 distinguishes the grass and non-grass areas according to the image collected by the image acquisition device 140, and then recognizes the grass boundary. Among them, non-grass areas include fences, sidewalks, low shrubs, road teeth, wood chips, etc.

The energy module 103 is used to provide energy for the operation of the mobile device 100. The energy source of the energy module 103 may be gasoline, a battery pack, etc. In this embodiment, the energy module 103 includes a rechargeable battery pack arranged in the housing 110. During operation, the battery pack releases electric energy to maintain the operation and movement of the mobile device 100. When not working, the battery can be connected to an external power source to supplement power. In particular, due to a more user-friendly design, when the battery power is insufficient, the self-mobile device 100 will automatically find a charging station to supplement the power.

The mobile device also includes a storage unit for storing a data model. The data model may contain a large number of different lawns, pictures information or other characteristic data of objects around the lawn, different houses, etc. The image acquisition device 140 collects peripheral images from the mobile device, and the control module 101 compares the collected images with the stored data model, analyzes the location and environment where the mobile device is located, and then controls the mobile device to move and move on the corresponding lawn. jobs.

Please refer to FIG. 2, the self-mobile device 100 includes an edge mode 108 and an edge detection mode 109. In the edge mode 108, the control module 101 controls the mobile device 100 to move and cut along the edge. In the edge detection mode, the control module 101 controls the self-mobile device 100 to move and cut within the boundary. The control module 101 is used to control the self-mobile device 100 to automatically switch between the edge mode and the edge detection mode. Specifically, the self-mobile device 100 can control the self-mobile device 100 to select the edge mode and the edge detection mode according to a preset time schedule. The schedule can be set at the factory, or it can be set by the user according to the user's usage habits, or it can be automatically set by the mobile device 100 according to the user's usage habits. Of course, you can also set the mode switch button, and the user can switch the working mode by himself.

In the edge mode 108, the image information of the target area is first acquired by the image acquisition device 140, and the control module 101 searches for the boundary according to the image information acquired by the image acquisition device, and controls the mobile device 100 to move and cut along the boundary. Specifically, the control module 101 first obtains image information from the image acquisition device, and then performs visual recognition based on the image information, image processing searches for boundary information, and after finding the boundary, controls the mobile device 100 to cut along the edge.

As shown in FIGS. 11 to 13, in the edge mode, the control module 101 is configured to determine whether the proportion of the non-working area and the working area in the environment image exceeds a threshold according to the environment image collected by the image collecting device; If it exceeds, perform the following steps:

Fitting the boundary of the working area in the environmental image into a fitting boundary line 31;

Generate a target point D according to the fitted boundary line 31;

Controlling the self-mobile device 100 to move to the target point D;

The self-moving device 100 is controlled to move along the boundary of the working area.

In an embodiment, before judging whether the proportion of the non-working area to the working area in the environmental image exceeds a threshold value according to the environment image collected by the image collecting device 140, the method further includes: according to the image collecting device 140 The collected environment image is analyzed whether there is a boundary of the working area in the environment image. When the working area boundary does not exist in the environment image, the moving module is controlled to move according to the preset edge finding logic to find the working area boundary; when the working area boundary exists in the environment image, according to The environment image collected by the image acquisition device is then executed to determine whether the proportion of the non-working area and the working area on both sides of the boundary in the environment image exceeds a threshold. If it is determined whether the proportion of the non-working area and the working area on both sides of the boundary in the environmental image exceeds the threshold, the following steps are performed:

Generate a target point D according to the fitted boundary line 31;

Controlling the self-mobile device 100 to move to the target point D;

If it is determined that the proportion of the non-working area and the working area on both sides of the boundary in the environmental image does not exceed the threshold, the self-mobile device is controlled to move according to a certain movement path, for example, the self-mobile device can be controlled to continue to move forward in the original direction until The proportion of the non-working area and the working area on both sides of the boundary in the environmental image exceeds the threshold.

In the foregoing embodiment, as shown in FIG. 11, a parallel line L separated by a predetermined distance from the fitting boundary line 31 is defined as a target line, wherein the target point D is any point on the target line L.

Specifically, generating a target point D according to the fitted boundary line 31 includes: selecting a point in the fitted boundary line 31 as the target base point F, and the line connecting the target point D and the target base point F It is perpendicular to the fitting boundary line 31, and the distance between the target point D and the target base point F is a preset distance. Wherein, the preset distance is equal to the distance from the fitting point E to the side 35 of the mobile device 100, or equal to the distance from the fitting point E to the side 35 of the mobile device 100 plus one preset distance. Set the safety distance.

As shown in Figures 11 to 13, a point on the self-mobile device is defined as the fitting point E of the self-mobile device. Controlling the movement of the self-mobile device to the target point D includes: controlling the movement of the self-mobile device so that the fitting point E of the self-mobile device moves to the target point; rotating the self-mobile device so that the driving direction of the self-mobile device is aligned with the fitted boundary line The extension direction is the same.

Wherein, the self-moving device 100 can move from the fitting point E to the target point D along an arbitrary path such as a curve or a straight line. In one embodiment, take the mobile device 100 moving from the fitting point E to the target point D along a straight line as an example. In this embodiment, the movement of the self-mobile device is controlled to move the fitting point of the self-mobile device to the target point. It includes: controlling the mobile device to move along the line connecting the fitting point and the target point, so that the fitting point is moved to the target point. Rotating the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line includes: according to the line between the fitting point and the target point and the fitting straight line Rotate the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line.

As shown in FIGS. 11 to 13, the image capture device 140 includes a side image capture device located on one side of the self-mobile device and a middle image capture device located in the middle of the self-mobile device. Among them, the side image acquisition device on one side is used to acquire image information in the edge edge mode, and may also be referred to as the edge image acquisition device 141. Wherein, the middle image acquisition device is used to acquire image information in the edge detection mode, and it may also be referred to as the edge detection image acquisition device 142. In the case of the side image acquisition device, the rotating the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line includes: rotating the self-moving device The equipment to the side where the side image acquisition device is located is close to the boundary of the working area.

After the rotating direction of the mobile device is the same as the extension direction of the fitted boundary line, the self-mobile device is already on the boundary of the working area. At this time, the self-mobile device is controlled to move along the boundary of the working area. Specifically, controlling the mobile device to move along the boundary of the working area includes:

Controlling the self-moving device to move along the current fitting boundary line. In the steps of fitting the environment image into a fitting boundary line in each of the foregoing embodiments, the entire environment image can be directly fitted into a fitting boundary line, and the fitting boundary line may be a straight line or a curve. It is also possible to divide the environment image into several sub-images, and then fit each sub-image into a straight line boundary. Specifically, in this embodiment, in the edge mode, the control module 101 is configured to analyze whether there is a working area boundary in the environment image according to the environment image collected by the image acquisition device; when there is no working area in the environment image When bordering, control the movement module to move according to the preset edge finding logic to find the boundary of the working area. When the working area boundary exists in the environment image, divide the environment image into N sub-images and analyze whether there is a working area boundary in each of the sub-images, and simulate the working area boundary in each sub-image with the working area boundary. Synthesize a straight line, which can also be called a straight line boundary line, and generate straight line parameters, and control the self-moving device to move and work along the working area boundary according to the parameters, where N≥2.

Generally, when processing environmental images, all boundary data points are usually connected end to end to fit a curve, or all boundary data points representing the boundary of the working area in the environmental image are directly fit to a straight line. If all the boundary data points are directly fitted into a curve, the calculation amount is too large, and the calculation capability of the control module 101 is very high, and it is time-consuming. Not only that, because there may be jumping and distorted boundary data points in the environment image. , If all data points are directly fitted into a curve, the movement path of the self-mobile device 100 may be distorted, and may even be severely distorted; moreover, if the curve is fitted, the movement trajectory of the self-mobile device is a curve. Since the movement of the mobile device requires constant adjustment of the direction, the movement posture is not good-looking. If all the boundary data points are fitted into a straight line, because the area spanned by an environmental image is large, after all the boundary points are fitted into a straight line, the data may be severely distorted, resulting in serious distortion of the last moving path from the mobile device. In this embodiment, the environment image is divided into N sub-images, and the boundary of the working area in each sub-image is respectively fitted into a straight line, and the number of sub-images is controlled at the same time, specifically N≥2, which improves the fitting The degree of realism of the boundary is reduced, the amount of calculation is reduced, the frequency of adjusting the direction from the mobile device is reduced, the moving posture is also more beautiful, it also reduces the requirements for the control module 101, reduces the cost, and when the sub-image is fitted with a straight line , Can filter out the distorted boundary data points, making the fitted boundary more accurate.

In this embodiment, the number of segmented sub-images is further limited, specifically, 2≤N≤8, and the number of sub-images is limited to 2 to 8 (including 2 and 8). After fitting, The accuracy of the boundary and the difficulty of calculation during fitting are controlled to the best, and the distorted boundary data points are effectively removed.

Specifically, the rules for dividing the environment image into N sub-images can be set according to actual conditions. For example, in terms of size, the environment image can be divided into N sub-images of the same size, or divided into N sub-images of different sizes. ; For another example, in the direction of division, it can be divided along the left and right direction, can also be divided along the up and down direction, can also be divided along the inner and outer directions, can also be divided along other rules and so on.

Taking the segmentation of the environment image along the distance from the mobile device 100 as an example, the region closest to the self-mobile device is obtained from the sub-image closest to the self-mobile device, and the self-mobile device is controlled according to the sub-image closest to the self-mobile device. The movement and work of the mobile device, and the situation of the area farther from the mobile device is obtained from the sub-images far away from the mobile device, so as to predict the subsequent movement and work of the mobile device.

For example, as shown in FIG. 4 and FIG. 7, the image acquisition device 140 can acquire an image of the target area M. On the image of the target area M, the front of the image is farther from the mobile device 100, and the rear of the image is self-moving. The device is close, so the image of the target area M is segmented along the front and back direction of the image of the target area M. Specifically, the control module 101 is further configured to: when there is a working area boundary in the environment image, divide the environment image into N sub-images along the direction of the distance from the mobile device, and divide the environment image into N sub-images according to the closest to the self-mobile device. The sub-image controls the movement and work of the self-mobile device, and predicts the subsequent movement and work of the self-mobile device based on the remaining sub-images. Specifically, when controlling the movement and work of the mobile device according to each sub-image, it is also firstly determined whether there is a working area boundary in each sub-image, and the working area boundary in each sub-image that has a working area boundary is respectively fitted into A straight line, and generate linear parameters; according to the parameters generated by the sub-image closest to the self-mobile device, control the self-mobile device to move and work along the current working area boundary; according to the parameters generated by the remaining sub-images, follow the self-mobile device Make predictions about your moves and work.

When the environment image is divided into N sub-images, the number of N can be controlled within 2 to 8. In this embodiment, as shown in Figures 7 to 9, in order to further reduce the difficulty of calculation, the environment image is moved along the distance. The device 100 is divided into two sub-images in the near and far direction, that is, it is divided into two sub-images along the front and back direction of the environment image. The sub-image in front of the environment image is farther from the mobile device, and the distance in the back of the environment image is Since the mobile device is closer. The control module 101 controls the current movement and work of the self-mobile device 100 according to the sub-images that are closer to the mobile device 100; predicts the subsequent movement and work of the self-mobile device 100 according to the sub-images that are farther from the mobile device 100 .

Specifically, in the edgewise mode, the control module 101 is further configured to: when there is a working area boundary in the environment image, divide the environment image into two sub-images along the distance from the self-mobile device, and move the image according to the proximity. The sub-image of the device controls the movement and work of the self-mobile device, and the subsequent movement and work of the self-mobile device are predicted based on the sub-images far away from the self-mobile device. Specifically, when controlling the movement and work of the mobile device according to the two sub-images, it is also first to determine whether there is a working area boundary in the two sub-images. If both of them exist, the working area boundary in each sub-image is respectively fitted. Form a straight line, and generate linear parameters; control the self-mobile device to move and work along the boundary of the current work area according to the parameters generated by the sub-image close to the self-mobile device; according to the parameters generated by the sub-image far away from the self-mobile device, control the self-mobile device to move and work along the boundary of the current work area; Make predictions about the subsequent movement and work of mobile devices. Of course, if there is no working area boundary in the sub-image close to the self-mobile device, the spatial relationship between the working area boundary and the self-mobile device can be predicted directly based on the parameters generated in the sub-image far away from the self-mobile device, and the self-mobile device can be controlled along Move the corresponding azimuth to find the boundary of the work area. If there is no working area boundary in the sub-image far away from the mobile device, but there is a working area boundary in the sub-image of the device close to the free land, you can first control the self-mobile device to control the self-mobile device according to the parameters generated by the sub-image close to it Move and work along the boundary of the working area, and predict the subsequent movement from the mobile device based on the information of the far away sub-image without the boundary of the working area, for example, plan the steering from the mobile device in advance.

As shown in FIGS. 7 to 9, it is taken as an example that the environment image includes the grass boundary, and the two sub-images both include the grass boundary. As shown in FIG. 8, it is a schematic diagram of an environmental image 300 taken from the mobile device 100 at a certain time, that is, a schematic diagram of an image taken in the target area M. The environmental image 300 includes grass 31 and non-grass 32. As shown in FIG. 9, it is a schematic diagram of dividing the environment image 300 into two sub-images along the distance from the mobile device 100 into two sub-images. The two sub-images are a near sub-image 301 and a far sub-image 302 respectively. The near sub-image 301 is the image of the near target area M1 closer to the mobile device in FIG. 7, and the far sub-image is the image of the far target area M2 farther from the mobile device in FIG. 7. The control module 101 further fits the borders of grass and non-grass in the two sub-images (301, 302) into a straight line respectively, wherein the borders of the grass 31 and the non-grass 32 in the near sub-image 302 are fitted to a near-boundary straight line 312 , And generate parameters representing the near-boundary straight line 312. The parameters may be the angle of the near-boundary straight line 312 and the offset of the near-boundary line relative to the center. Of course, it may also be other parameters that represent the near-boundary straight line 312. The boundary between the grass 31 and the non-grass 32 in the far sub-image 301 is fitted into a far boundary straight line 311, and a parameter representing the far boundary straight line 311 is also generated. The control module 101 controls the movement and operation of the self-mobile device along the nearby boundary line according to the parameters of the near-boundary straight line 312, and predicts the subsequent movement and work of the self-mobile device according to the parameters of the far-boundary straight line 311.

The above-mentioned method of dividing the environment image into several sub-images, and then respectively fitting the several sub-images into a straight boundary line, can be applied to any step of fitting the environment image into the fitting boundary line in the edge mode. Taking the step of controlling the movement of the self-mobile device along the boundary of the working area as an example, fitting the boundary of the working area in the current environment image to the current fitting boundary line includes:

Among them, 2≤N≤8 can be selected, and the number of sub-images is limited to 2 to 8 (including 2 and 8), which can control the accuracy of the boundary after fitting and the difficulty of calculation during fitting to the best , And effectively remove distorted boundary data points. In one embodiment, N=2, that is, the environment image is divided into 2 sub-images along the distance from the mobile device.

Of course, the above method of segmenting the environment image and then fitting a straight line can also be applied not only in the step of controlling the self-mobile device along the boundary of the working area, but in the early stage as shown in Figure 11, according to the fitting of the environment image fitting The boundary line, controlled from the mobile device, can also be applied in a step where the moving direction is the same as the extension direction of the boundary line. I will not repeat them in this application.

In the edge mode, the control module 101 controls the mobile device to move and work along the boundary of the working area according to the parameters obtained from each sub-image, and judges whether the boundary of the working area is lost in real time. Specifically, in the edge mode, the control module 101 is further configured to determine in real time whether the current working area boundary is lost, and when the current working area boundary is lost, control the moving module to automatically move to find the Work area boundary; when the current work area boundary is not lost, control the self-mobile device to continue to move and work along the current work area boundary.

Wherein, determining whether the current working area boundary is lost in real time includes: by comparing the ratio of the non-working area and the working area on both sides of the current working area boundary, and by comparing the non-working area and the working area on both sides of the current working area boundary. When the ratio of the non-working area to the working area is within the preset range, it is determined that the boundary of the current working area is not lost; when the ratio of the non-working area to the working area is not within the preset range When it is time, it is judged that the boundary of the current working area is lost. That is, the proportion of non-standard objects and target objects is counted. When the ratio of non-standard objects to target objects is within the preset range, it is judged that the current working area boundary is not lost; when the ratio of non-standard objects to target objects is not within the preset range At the time, it is judged that the boundary of the current working area is lost. In this embodiment, the working area is the grass boundary, and grass can be selected as the target object. The ratio of non-grass to grass on both sides of the current grass boundary can be counted to determine whether the current grass boundary is lost. Specifically, in the edge-edge mode, the control module 101 is further configured to: perform statistics on the ratio of non-grass to grass on both sides of the current grass boundary in real time, and when the ratio of non-grass to grass is within a preset range, Determine that the current grass boundary is not lost, and control the mobile device to continue moving and mowing along the current grass boundary; when the ratio of non-grass to grass is not within the preset range, determine that the current grass boundary is lost, and control the mobile module to automatically move to find the grass boundary . Of course, the above is only a method for judging whether the boundary of the working area is lost, and in other embodiments, it can also be judged by other methods.

Wherein, when the boundary of the working area is lost, controlling the moving module to automatically move to find the boundary of the working area may specifically include: when the boundary of the current working area is lost, controlling the moving module to rotate a certain angle to find the boundary of the working area; The boundary of the working area is controlled to move the moving module according to the preset edge finding logic to find the boundary of the working area. Specifically, when the mobile module is controlled to move according to the preset edge finding logic, the mobile module can be controlled to search within the range of the two self-mobile device fuselages inside the working area. If the boundary is still not found, then within a larger range Look for. Of course, the above is only a searching method after the boundary of the working area is lost. In other embodiments, the boundary of the working area can also be found by other methods.

In the edge detection mode, the control module 101 controls the mobile device to move and cut grass within the boundary of the work area according to the information in the environment image. Specifically, the control module 101 is configured to: control the image acquisition device acquisition environment Image, and judge whether there is a working area boundary in the environment image according to the environment image; when there is no working area boundary in the environment image, control the movement module to move and work according to the preset edge movement logic; when there is a working area boundary in the environment image , Fit the boundary of the working area in the environmental image into a fitted boundary line. Specifically, the fitted boundary line can be a straight line or a curve, and the parameters of the fitted boundary line are generated, and the self-mobile device is controlled to work according to the parameters. Move and work within the boundaries of the area. In the edge detection mode, the self-mobile device 100 can cut randomly within the boundary of the working area, or cut along a planned path, for example, using inertial navigation and an odometer to make an I-shaped path for cutting.

As shown in Figure 4, because the environmental image captured by the image capture device has a blind spot, the blind spot distance is A, and the image capture device cannot capture the area where the distance from the mobile device is A on the ground. Therefore, when the image capture device captures When there is a grass boundary in the environment image of, it does not mean that the self-mobile device has encountered the grass boundary. At this time, the distance between the self-mobile device and the grass boundary is at least A. For example, as shown in Figure 4, when calculating from the environment image When the distance from the mobile device to the grass boundary is B, at this time, the actual distance from the mobile device to the grass boundary is B+A. Therefore, in order to allow the self-mobile device to better cut to the edge, when the self-mobile device is calculated based on the environment image that the self-mobile device is close to the boundary of the working area, the control module 101 controls the self-mobile device to continue to move a certain distance in the original moving direction. The certain distance can be directly preset at the time of leaving the factory. For example, the preset distance is less than or equal to the above-mentioned blind zone distance A, so that the self-mobile device can better cut to the edge in the working area. Of course, the certain distance can also be directly generated by the current calculation, or obtained by other means. Specifically, in the edge detection mode, when there is a working area boundary in the environmental image, controlling the self-moving device to move and work within the working area boundary according to the parameter includes: when judging self-moving according to the environmental image When the device is close to the boundary of the working area, control the self-moving device to continue to move a preset distance in the original direction, and then adjust the moving direction of the self-moving device.

As shown in Figure 10, the work area includes at least two sub-work areas. The control module 101 further determines whether the current sub-work area is cut. If the cut is completed, it enters the next sub-work area. If the cut is not completed, it continues to work in the next sub-work area. Cutting on the sub-work area. Since the self-mobile device has the edge mode and the edge detection mode, the control module 101 can control the self-mobile device to complete the work tasks on the boundary of each sub-work area in the edge mode, and complete the tasks in each sub-work area in sequence in the edge detection mode. Task. Alternatively, first switch the edge mode and edge detection mode to complete the work tasks on and within the boundary of a sub-work area, and then enter the next sub-work area, switch the edge mode and edge detection mode to complete the next sub-work area. Work tasks on and within the boundaries of the work area.

In order to complete the edge work tasks and the work tasks within the boundaries of the respective work areas in sequence, and in each sub-work area, first execute the edge mode to perform the work tasks at the boundary of the work area, and then execute the edge detection mode to perform the work within the work area. Take the work task as an example. The control module 101 judges whether the cutting of the working area boundary is completed in real time or at intervals. If the cutting is not completed, the cutting continues. If the current boundary cutting is completed, the mobile device 100 is controlled to enter the edge detection mode to cut the The area within the boundary of the sub-work area; when the area within the boundary of the sub-work area is cut, enter the next sub-work area to work.

The specific method for judging whether the work task on the work area is completed can be achieved by setting a reference object, for example, laying a magnetic stripe on the boundary of the work area, or by setting a positioning device on a self-mobile device.

As shown in FIG. 10, the work area includes at least two sub-work areas. The mobile device also includes a reference object detection module for detecting a reference object. The control module 101 controls the self-mobile device to automatically switch at least according to the information detected by the reference object detection module. Move and cut within two sub-work areas. Specifically, when the self-mobile device finishes working in a sub-work area, the control module 101 controls the self-mobile device to search for the reference object along the edge according to the environmental image collected by the image acquisition device, and when the reference object detection module detects When referring to the reference object, the control module 101 controls the self-mobile device to enter the next sub-work area according to the information detected by the reference object detection module. The completion of the above-mentioned sub-work area work can be in a single edge mode, or in the edge detection mode, and the sub-work area work is completed, but also in the edge mode and the edge detection mode, the work in the sub-work area is completed.

As shown in FIG. 10, the automatic working system includes at least two sub-work areas. In this embodiment, each sub-work area is called a first sub-area 11, a second sub-area 12, a third sub-area 13, and a fourth sub-area. 14. The automatic working system also includes a number of magnetic strips 15 for connecting at least two sub-regions. The magnetic strip 15 includes a first magnetic strip 151 for connecting the first sub-region 11 and the second sub-region 12, and a first magnetic strip 151 for connecting the second sub-region. 12 and the second magnetic stripe 152 of the third subarea 13 are used to connect the fourth magnetic stripe 153 of the third subarea 13 and the fourth subarea 14. The control module 101 judges whether the self-mobile device 100 cuts the corresponding boundary according to the magnetic stripe 15. For example, the control module 101 controls the self-mobile device 100 to start cutting from the first magnetic stripe 151, and when it detects the first magnetic stripe again At 151, it indicates that the boundary cutting of the first sub-region 11 is completed, and the other sub-regions can also be tested whether the cutting is completed by the same method.

After the boundary cutting of the first sub-region is completed, the mobile device 100 can directly move through the first magnetic strip 151 from the boundary of the first sub-region 11 to the boundary of the second sub-region 12 to complete the second sub-region 12 The work of the boundary of, in the same way, moves from the boundary of the second sub-region 12 to the boundary of the third sub-region 13 through the second magnetic stripe 152, and moves from the boundary of the third sub-region 13 through the third magnetic stripe 153 To the boundary of the fourth sub-region 14. Of course, in other embodiments, the self-mobile device 100 can also detect whether the boundary has been cut and switch between the two sub-regions in other ways, for example, by identifying with a QR code or other marks or by using the self-mobile device 100 Set up positioning devices and other methods.

Of course, after the boundary cutting of the first sub-region is completed, the mobile device 100 may first cut the boundary of the first sub-region, and after the cutting in the boundary of the first sub-region is completed, it will search for the magnetic stripe along the edge and pass the first magnetic The bar 151 enters the boundary of the second sub-work area 12 to sequentially complete the work on and within the boundary of the second sub-work area 12, and similarly, complete the work of other sub-work areas in turn. In each sub-work area, the order of execution on and within the boundary of the work area can be set according to actual conditions. For example, the work within the boundary of the work area can also be executed first, and then the work on the boundary of the work area can be executed.

Specifically, in the edge detection mode and the edge detection mode, the specific means for processing the environmental image to analyze whether there is a boundary of the working area in the environmental image can be selected according to the actual situation. For example, in this embodiment, in the edge mode, the mobile device 100 can sequentially process the environment image by means of distortion correction, image segmentation, perspective transformation, etc., to generate several data points representing the boundary of the working area, and then the environment image Divide into N sub-images, and then fit the boundary data points in each sub-image into a straight line, calculate the parameters representing the straight line, for example, calculate the angle and offset of the straight line, and then control according to the parameters Move and work from mobile devices. Of course, in the edge mode, the environment image can also be directly divided into N sub-images, and then each sub-image is processed by means of distortion correction, image segmentation, perspective transformation, straight line fitting, etc., to generate each fitted straight line Parameters. Of course, the above-mentioned processing method for the environmental image is only an example, and other processing methods in the field may also be used.

As shown in Figure 4, the image capture device is installed on the mobile device 100, the installation angle of the image capture device is 70-150 degrees, and the installation height (distance from the ground) H is 10-40cm (for example, the installation height is 14-15cm or 20cm, Or 30cm, etc.), the angle α is 20-90 degrees, where the distance D refers to the distance seen by the image capture device, or in other words, the distance of the image that the image capture device can capture. The image acquisition device may be one or two or more.

As shown in Figures 5 and 6, in an embodiment, the self-mobile device includes at least two image acquisition devices 140, and the two image acquisition devices 140 are respectively an edge image acquisition device 141 and an edge image acquisition device 141 for acquiring image information in the edge mode. An edge-detecting image acquisition device 142 for acquiring image information in the edge-detecting mode. In the edge mode, the control module 101 controls the movement and cutting of the self-mobile device according to the environmental image collected by the edge image acquisition device; in the edge detection mode, the control module 101 controls the self-movement according to the environmental image collected by the edge image acquisition device Moving and cutting of equipment.

As shown in FIG. 5, in the edgewise mode, the moving direction of the self-mobile device 100 is parallel to the extension direction of the boundary, that is, the self-mobile device 100 moves along the edge, because the edge-edge image capture device 141 has a limited viewing angle, and in order to make the image capture device 141 can collect border images in real time to avoid the loss of border images. The border image acquisition device 141 is installed on the side of the self-mobile device 100 close to the border. In order to avoid the front view of the border image acquisition device 141 from being blocked, it is generally set in the self-moving device. The front of the moving direction of the device 1 and the side close to the boundary. Specifically, the edge image acquisition device 141 can be arranged in the front of the moving direction of the self-mobile device, and within the range of the distance S1 from the side of the self-mobile device close to the boundary, S1 can be determined by the field of view of the edge image acquisition device. For example, in this embodiment, S1 is 0-5 cm. Of course, in other embodiments, S1 can also select other ranges according to actual conditions. In the height direction, the installation height of the edge image acquisition device shall not exceed 20cm above the ground.

As shown in FIG. 6, in the edge detection mode, the edge detection image acquisition device 142 collects image information of the environment in which the boundary is moved from the mobile device 100, and the control module 101 controls the mobile device 100 in Move and cut within the boundary. The edge detection image acquisition device 142 is located in front of the moving direction of the self-mobile device. As shown in FIG. 6, in the edge detection mode, the self-mobile device 100 moves perpendicular to the line where the boundary is, that is, the self-mobile device 100 moves along the line perpendicular to the boundary. When it moves to the edge detection image acquisition device 142 to collect the boundary image, it continues to move forward by a preset distance L, and then turns around to prevent the mobile device 100 from driving out of the boundary. The edge detection image acquisition device 142 is set within the range of the distance S2 from the central axis of the mobile device. S2 can be determined by the field of view of the edge image acquisition device to ensure that the edge detection image acquisition device can collect in real time when the mobile device 100 moves toward the boundary. To the border image. For example, S2 in this embodiment is 0-4 cm. Of course, in other embodiments, S2 can also be selected in other ranges according to actual conditions. In the height direction, the installation height of the edge detection image acquisition device shall not exceed 20cm above the ground.

In summary, as shown in Figures 5 and 6, in the left and right directions of the self-moving device, the edge image acquisition device is set on the side of the self-moving device close to the working area boundary, and the edge detection image acquisition device is set on the self-moving device The side close to the center of the self-mobile device; in the front and back direction, the edge-edge image acquisition device and the edge-detection image acquisition device are both set in front of the forward direction of the self-mobile device; in the height direction, the edge-edge image acquisition device and the edge-detection image acquisition device The installation height does not exceed 20cm above the ground; in the left and right direction, the distance between the edge image acquisition device and the side of the mobile device close to the boundary of the working area is S1, S1 is within the range of 0-5cm, the edge detection image acquisition device The distance from the central axis of the mobile device is S2, and S2 is in the range of 0-4 cm.

Of course, in another embodiment, there may be only one image acquisition device. For example, an image acquisition device that can be moved is provided. When the self-mobile device 100 is in the edge mode, the image acquisition device is in the first state. When the device 100 is in the edge detection mode, the self-mobile device is in a second state different from the first state, so that it is adapted to the image acquisition angle required by the mode it is in.

Or, when the installation range of the edge detection image acquisition device 142 and the installation range of the edge image acquisition device 141 can overlap, only one image acquisition device may be provided. The image acquisition device is set in the overlapping area, and the image acquisition angle is It can meet the requirements of edge detection mode and also meet the requirements of edge mode. For example, in one embodiment, when the image capture device is set at a height of 30 cm or more from the ground, and the distance from the side of the mobile device close to the boundary is S0 (for example, S0 is 1/4 to 1/ of the width of the mobile device). 3) The image acquisition device can serve as both an edge image acquisition device and an edge detection image acquisition device. Of course, the above are only distances. In other embodiments, other numbers of image acquisition devices 140 may also be provided.

In this embodiment, the environment image can also be collected by the above-mentioned image acquisition device, and the control module judges whether there is an obstacle in front according to the environment image collected by the image acquisition device, and controls the mobile device to automatically avoid the obstacle.

Further, the control module obtains the obstacle type in the environmental image according to the environmental image collected by the image acquisition device, and controls the mobile device to execute an action corresponding to the current obstacle type. Specifically, the control module can automatically recognize the type of obstacle ahead in the environment image, and control the mobile device to execute an action corresponding to the current obstacle type according to the recognized different obstacle types. After collecting the environmental image, it can also send the image to the cloud, calculate in the cloud, identify the obstacle type, and send the recognition result to the mobile device. The control module obtains the recognition result and controls the execution of the mobile device and the current obstacle. The action corresponding to the type. In an embodiment, the types of obstacles may include humans, animals, accessible obstacles, and inaccessible obstacles, among which animals include cats, dogs, hedgehogs, etc.; accessible obstacles include houses, trees, flower beds, fences, and roads. Teeth, etc.; inaccessible obstacles include ponds, roads, etc.

When the control module obtains the recognition result that the obstacle in front is a person, the control module controls the mobile device to turn or turn around to avoid the person, or stop motion to avoid collision with the person; of course, the self-mobile device can also first recognize the person ahead. Reduce the speed to avoid collisions with people due to fast speed, turning or turning around. In one embodiment, when it is recognized that the obstacle in front is a person, the self-moving device can also interact with the person, specifically through voice interaction, For example, warning voices such as "Danger, please avoid" can be played; of course, it can also be interacted in other ways, such as flashing warning lights. After interacting with the person, if it is recognized that the person has left, the self-mobile device is controlled to continue forward, and if it is recognized that the person has not left, the self-mobile device is controlled to turn or turn around or stop moving to avoid collision with the person.

When the control module obtains the recognition result that the obstacle in front is an animal, for example, when it recognizes that there are cats, dogs, hedgehogs, etc., it can control the mobile device to slow down, and/or send warning voices or lights to drive away. If the repelling of the animal is invalid, control the self-moving device to turn or turn around to avoid the animal. Of course, when the non-contact drive such as slowing down, voice or light is invalid, the self-mobile device can be further controlled to collide with the front obstacle at low speed to further drive the animal. After driving the animal in the above manner, if it is recognized that the animal has left, the self-mobile device is controlled to continue forward, and if it is recognized that the animal has not left, the self-mobile device is controlled to turn or turn around to avoid the animal. In some embodiments, to further ensure animal safety, the self-moving device further includes a collision sensor. When the self-mobile device collides with an obstacle in front, the collision sensor detects the collision, and the control module controls the self-mobile device to turn or turn around to avoid the animal.

When the control module obtains the recognition result that the front obstacle is a contactable obstacle, for example, when it recognizes that there are houses, trees, flower beds, etc., the control module controls the mobile device to approach the front obstacle as much as possible. The cutting area is completely cut. When approaching the accessible obstacle, turn or turn around to avoid the obstacle in front. In order to further ensure safety, when it is recognized that there is a house, tree, flower bed and other contactable obstacles ahead, first reduce the speed to avoid too fast and hit the obstacle in front, and drive at a reduced speed until it is close to the obstacle in front and then turn or turn around. Avoid obstacles ahead. In order to cut the self-moving device as close to the accessible obstacle as possible to achieve complete cutting of the cutable area, in some embodiments, the self-moving device also includes a collision sensor. When the self-mobile device recognizes that there are houses, trees, When the flower bed or other contactable obstacles, control the mobile device to slow down and move forward at low speed until it collides with the obstacle in front. The collision sensor detects the collision, and the control module controls the mobile device to turn or turn around. On the one hand, try to move as close to the obstacle as possible The area is completely cut; on the other hand, the low-speed collision ensures that the accessible obstacles are not damaged.

When the control module obtains the recognition result that the front obstacle is an inaccessible obstacle, for example, when it recognizes that there is a pond or a border in front, the control module controls the mobile device to drive to approach the front obstacle, and when it approaches the front obstacle, Turn or turn around to avoid contact with inaccessible obstacles, for example, to avoid rushing into a pond or rushing onto the road. In order to further ensure safety, when an inaccessible obstacle such as a pond or a road is identified in front, first slow down and drive until it is close to the obstacle in front, and then turn or turn around to avoid contact with inaccessible obstacles, avoid too fast, and rush into Or hit the obstacle ahead.

In the foregoing embodiment in which the control module controls the self-mobile device to adopt different motion strategies for different front obstacles, when the control module controls the self-mobile device to approach the front obstacle, the blind spot distance as shown in Figure 4 should be considered. A, so that the self-moving device can cut as close to the obstacle as possible to completely cut the cutable lawn area.

In the above process of recognizing obstacles in the environmental image, the recognition can be performed by means of image detection, image segmentation, and image classification on the environmental image to obtain the obstacle type in the environmental image. The above-mentioned image detection, image segmentation, image classification and other methods can automatically extract features based on neural network methods. Of course, features can also be extracted by traditional methods of manually setting features. The model of the neural network can obtain the recognition result by computing from the mobile device, or deploy the model in the cloud, and obtain the recognition result through cloud computing.

In this embodiment, only the above four different obstacle types are used as examples to illustrate their corresponding motion strategies. In other embodiments, the control module may also set other obstacle types and corresponding motion strategies, or combine the above four types of obstacles. The object type matches other sports strategies. This application will not give examples one by one.

The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the various technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, All should be considered as the scope of this specification.

The above-mentioned embodiments only express several implementation modes of the present invention, and their description is relatively specific and detailed, but they should not be understood as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims

A self-moving device, characterized in that it comprises:

case;

The moving module is located below the housing and is used to drive the housing to move;

The working module is arranged on the housing to perform preset working tasks;

An image acquisition device for acquiring an image of the environment in which the self-mobile device is located;

The control module is configured to autonomously control the movement module to drive the housing to move, and autonomously control the working module to perform preset working tasks;

The self-moving device includes an edge mode, and in the edge mode, the control module is configured to:

According to the environment image collected by the image acquisition device, determine whether the proportion of the non-working area to the working area in the environment image exceeds the threshold; if it exceeds, then:

Fitting the boundary of the working area in the environmental image into a fitting boundary line;

Generating a target point according to the fitted boundary line;

Controlling the self-mobile device to move to the target point;

Controlling the self-moving device to move along the boundary of the work area.
The self-moving device according to claim 1, wherein a parallel line separated by a preset distance from the fitted boundary line is defined as a target line, and the target point is a point on the target line.
The self-mobile device according to claim 2, wherein said generating a target point according to said fitted boundary line comprises: selecting a point in said fitted boundary line as a target base point, and said target point The connecting line with the target base point is perpendicular to the fitting boundary line, and the distance between the target point and the target base point is a preset distance.
The self-mobile device according to claim 3, wherein: defining a point on the self-mobile device as a fitting point of the self-mobile device, and said controlling the self-mobile device to move to the target point comprises: The movement of the self-mobile device is controlled so that the fitting point of the self-mobile device moves to the target point.
The self-moving device according to claim 4, wherein the preset distance is equal to the distance from the fitting point to one side of the self-mobile device, or equal to the distance from the fitting point to the self-moving device. Add a safety distance to the distance on one side of the device.
The self-moving device according to claim 4, wherein said controlling said self-moving device to move to said target point further comprises:

Rotate the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line.
7. The self-moving device according to claim 6, wherein controlling the self-moving device to move so that the fitting point of the self-moving device moves to the target point comprises: controlling the self-moving device, Moving along the line connecting the fitting point and the target point, so that the fitting point moves to the target point.
7. The self-moving device according to claim 7, wherein the rotating the self-moving device so that the driving direction of the self-moving device is the same as the extension direction of the fitted boundary line comprises: according to the The angle between the line connecting the fitting point and the target point and the fitting straight line rotates the self-moving device so that the traveling direction of the self-moving device is the same as the extension direction of the fitting boundary line.
8. The self-moving device according to claim 8, wherein the image acquisition device comprises a side image acquisition device located on one side of the self-moving device, and the rotating the self-moving device makes the self-moving device The traveling direction of the mobile device is the same as the extension direction of the fitted boundary line, including: rotating the self-mobile device to a side where the side image acquisition device is located close to the working area boundary.
The self-moving device according to claim 1, wherein the controlling the self-moving device to move along the boundary of the working area comprises:

Controlling the self-mobile device to move forward and synchronously collecting current environment images;

Fitting the boundary of the working area in the current environment image to the current fitting boundary line;

Controlling the self-moving device to move along the current fitting boundary line.
The self-mobile device according to claim 10, wherein the fitting the boundary of the working area in the current environment image to the current fitting boundary line comprises:

Divide the current environment image into N sub-images along the distance from the mobile device, where N≥2;

Fitting the boundary of the working area in each of the sub-images into a straight boundary line;

Control the self-moving device to move along the linear boundary line synthesized by the sub-image closest to the self-mobile device, and the linear boundary line synthesized according to the remaining sub-images is the boundary line of the self-mobile device Make predictions for subsequent sports.
The self-moving device of claim 11, wherein N=2.
The self-moving device according to claim 1, wherein the control module is configured to determine the difference between the non-working area and the working area in the environment image according to the environment image collected by the image collecting device. Before the proportion exceeds the threshold, the method further includes: analyzing whether there is a boundary of the working area in the environment image according to the environment image collected by the image acquisition device, and when the working area does not exist in the environment image When the boundary is reached, the movement module is controlled to move according to the preset edge finding logic to find the boundary of the working area; when the boundary of the working area exists in the environment image, according to the environment image collected by the image acquisition device, It is determined whether the proportion of the non-working area and the working area on both sides of the boundary in the environmental image exceeds a threshold.
The self-moving device according to claim 1 or 13, wherein the control module is configured to:

If it is determined that the proportion of the non-working area to the working area in the environmental image does not exceed the threshold, the self-moving device is controlled to continue to move forward in the original direction.
The self-mobile device according to claim 1, wherein in the edge mode, the control module is further configured to: determine in real time whether the current working area boundary is lost, and when the current working area boundary is lost, The mobile module is controlled to move automatically to find the boundary of the working area; when the boundary of the current working area is not lost, the self-mobile device is controlled to continue to move and work along the boundary of the current working area.
The self-mobile device according to claim 15, wherein the real-time determination of whether the boundary of the current working area is lost comprises: performing a calculation of the ratio of the non-working area to the working area on both sides of the boundary of the current working area. According to statistics, when the ratio of the non-working area to the working area is within the preset range, it is judged that the current working area boundary is not lost; when the ratio of the non-working area to the working area is not within the preset range, it is judged The current working area boundary is lost.
The self-moving device according to claim 16, wherein when the current working area boundary is lost, controlling the moving module to automatically move to find the working area boundary comprises: when the current working area boundary is lost, controlling The moving module rotates a certain angle to find the boundary of the working area; if the rotation does not find the boundary of the working area, the moving module is controlled to move according to a preset edge finding logic to find the boundary of the working area.
The self-moving device according to claim 1, wherein the self-moving device further comprises an edge detection mode, and in the edge detection mode, the control module is configured to: control the collection of the image acquisition device The environment image, and determine whether the working area boundary exists in the environment image according to the environment image; when the working area boundary does not exist in the environment image, control the movement module to move within a preset edge Logical movement and work; when the working area boundary exists in the environment image, the working area boundary in the environment image is fitted into a fitting boundary line, and the parameters of the fitting boundary line are generated , And control the self-moving device to move and work within the boundary of the working area according to the parameter.
The self-moving device according to claim 18, wherein in the edge detection mode, when the working area boundary exists in the environmental image, the self-moving device is controlled according to the parameter in the Moving and working within the boundary of the working area includes: when judging from the environment image that the self-mobile device is close to the boundary of the working area, controlling the self-mobile device to continue to move a certain distance in the original direction, and then adjust the moving direction of the self-mobile device.
The self-moving device according to claim 1, wherein the self-moving device is an automatic lawn mower that automatically moves and cuts grass on the grass, and the working module is a mowing module for performing mowing tasks , The boundary of the working area is the boundary of the grass.
The self-moving device according to claim 1, wherein the image acquisition device comprises an edge image acquisition device and an edge detection image acquisition device, and in the edge mode, the control module is based on the edge image acquisition device The collected environment image controls the movement and cutting of the self-moving device; in the edge detection mode, the control module controls the self-movement according to the environment image collected by the edge detection image acquisition device Moving and cutting of equipment.
22. The self-moving device according to claim 21, wherein, in the left-right direction of the self-moving device, the edge edge image acquisition device is arranged on the side of the self-moving device close to the boundary of the working area, The edge detection image acquisition device is arranged on the side of the self-moving equipment close to the center of the self-moving equipment.
The self-moving device according to claim 21, wherein, in the front-to-rear direction, the edge edge image acquisition device and the edge detection image acquisition device are both arranged in front of the forward direction of the self-moving device; in the height direction Above, the installation height of the edge image acquisition device and the edge detection image acquisition device does not exceed 20 cm above the ground; in the left and right directions, the edge image acquisition device is close to the working area from the self-moving device The distance on one side of the boundary is S1, S1 is in the range of 0-5 cm, the distance between the edge detection image acquisition device and the central axis of the self-moving device is S2, and S2 is in the range of 0-4 cm.
The self-moving device according to claim 1, wherein the self-moving device further comprises a reference object detection module for detecting a reference object, and the control module controls the reference object detection module according to the information detected by the reference object detection module. The description is that the mobile device automatically moves and cuts in at least two sub-work areas.
The self-moving device according to claim 24, wherein when the self-moving device finishes working in a sub-work area, the control module controls the self-moving device according to the environmental image collected by the image acquisition device , Controlling the self-mobile device to find the reference object along the edge, and when the reference object detection module detects the reference object, the control module controls the self-mobile device to enter according to the information detected by the reference object detection module The next working sub-area.