Autonomous navigation method and system for a mobile sentinel robot

Abstract

The present invention discloses an autonomous navigation method and system of a mobile sentinel robot, including: obtaining a destination to be reached, setting navigation on a map, and advancing according to the navigation; obtaining road surface information of the advancing direction in real time, and judging whether the road surface information is passable by image processing, and obtaining portrait information in the road surface information if it is not passable; if there is portrait information in the road surface information, sending a set broadcast voice, receiving a reply voice, and converting the reply voice into text and extracting the reply meaning of the reply voice using a semantic extraction algorithm; correcting the original navigation according to the reply meaning to obtain a new navigation, and advancing according to the new navigation, and ending the operation when the destination is reached. The present invention obtains navigation according to a map, communicates with construction personnel at the same time, recognizes the voice of the construction personnel to extract the meaning therein, and corrects the navigation according to the meaning, and finally moves along with the navigation, thereby improving the efficiency of reaching the destination.

Description

Technical Field

[0001] The invention relates to the technical field of movable sentinel robots for construction, and in particular to an autonomous navigation method and system for the movable sentinel robot. Background Art

[0002] The mobile sentinel robot is an important helper in current construction, and plays a role in construction assistance at the construction site. Generally, the personnel inspection, project management, and real-time project follow-up required at the construction site all require the help of the mobile sentinel robot.

[0003] At present, the mobile sentinel robot shuttles through the ground of the construction site by navigation. Because of the rolling movement of the wheels at the bottom, it cannot pass through some areas with complex road conditions. In the construction site, a large amount of construction materials and tools are often randomly piled up on the construction site for the convenience of construction, which will cause the mobile sentinel robot to be unable to pass normally on the construction road according to the navigation map. Summary of the invention

[0004] The technical problem to be solved by the present invention is to address the deficiencies in the above-mentioned prior art and to provide an autonomous navigation method and system for a mobile sentinel robot, which obtains navigation according to a map, communicates with construction workers at the same time, recognizes the construction workers' voices to extract the meaning, and corrects the navigation according to the meaning, and finally moves according to the navigation, thereby improving the efficiency of reaching the destination.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: an autonomous navigation method of a mobile sentinel robot, comprising:

[0006] S1: Get the destination to be reached, set the navigation on the map, and move forward according to the navigation;

[0007] S2: Acquire the road surface information in the forward direction in real time, and determine whether the road surface information is passable by image processing. If passable, proceed according to the road surface information and repeat step S2. If not passable, obtain the portrait information in the road surface information and execute step S3.

[0008] S3: If the road information contains portrait information, a set broadcast voice is issued, and a reply voice is received. The reply voice is converted into text and then the meaning of the reply voice is extracted using a semantic extraction algorithm;

[0009] S4: Modify the original navigation according to the reply meaning to obtain a new navigation, and proceed according to the new navigation. If the destination is not reached, jump to step S2. If the destination is reached, end the operation.

[0010] Further, in step S2, the road surface information in the forward direction is obtained, including:

[0011] Get the current position of the mobile sentry robot;

[0012] Obtaining the position of the mobile sentry robot at the next time node according to the navigation;

[0013] Determine the forward direction according to the position of the mobile sentinel robot at the next time node;

[0014] Obtain road surface information based on the direction of travel.

[0015] Further, in step S2, determining whether the road surface information is passable is performed by image processing, including:

[0016] Converting the road surface information into a road surface image, and extracting roads in the road surface image;

[0017] Determine whether there is an obstacle on the road in the road image, and determine the actual size and position of the obstacle;

[0018] Determine whether the road surface is passable based on the actual size and position of the obstacle.

[0019] Furthermore, the determining whether there is an obstacle on the road in the road surface image includes:

[0020] Extracting the color value of each pixel in the road surface image, and comparing each color value with the color value in a set range;

[0021] Filter out the pixel points corresponding to the color values ​​within the set range;

[0022] Extract pixels that are clustered together from the selected pixels through a clustering algorithm;

[0023] Determine whether the pixels gathered together vertically penetrate the road image. If so, there is no obstacle. If not, there is an obstacle.

[0024] Furthermore, the determining of the actual size and position of the obstacle includes:

[0025] The obstacle contours in the road surface image are extracted by using a graphic contour extraction algorithm;

[0026] Obtain the size and position of obstacle outlines in the road image;

[0027] The actual size and position of the obstacle are calculated according to the focal length of the road surface image and the size and position of the obstacle in the road surface image.

[0028] Furthermore, the reply meaning includes an equal number of reply times and reply directions, and each of the reply times and the reply directions corresponds to each other one by one.

[0029] Furthermore, in step S4, the original navigation is modified according to the reply meaning to obtain a new navigation, including:

[0030] The remaining navigation after the current position in the original navigation;

[0031] A new remaining navigation is set according to the response time and the response direction, and the correction of the original navigation is completed to obtain the new navigation.

[0032] Based on the same inventive concept, the present invention also proposes an autonomous navigation system for a mobile sentinel robot, comprising:

[0033] The navigation generation module is used to obtain the destination to be reached, set the navigation on the map, and proceed according to the navigation;

[0034] The road surface judgment module is used to obtain the road surface information of the forward direction in real time, and judge whether the road surface information is passable by means of image processing. If it is passable, the vehicle moves forward according to the road surface information and repeats the road surface judgment module. If it is not passable, the vehicle obtains the portrait information in the road surface information and executes the voice inquiry module.

[0035] The voice inquiry module is used to send out a set broadcast voice if there is portrait information in the road information, and receive a reply voice, and then use a semantic extraction algorithm to extract the reply meaning of the reply voice after converting the reply voice into text;

[0036] The navigation update module is used to correct the original navigation according to the reply meaning to obtain a new navigation, and move forward according to the new navigation. When the destination is not reached, it jumps to the road surface judgment module, and when the destination is reached, the operation ends.

[0037] Compared with the prior art, the present invention has the following advantages:

[0038] The present invention provides an autonomous navigation method and system for a mobile sentinel robot, which obtains navigation according to a map, communicates with construction workers at the same time, recognizes the construction workers' voices to extract meanings, and corrects the navigation according to the meanings, and finally moves along the navigation, thereby improving the efficiency of reaching the destination. When the mobile sentinel robot encounters obstacles while walking on the road surface of a construction site, reasonable obstacle avoidance can be carried out through effective communication with the construction workers, so that the mobile sentinel robot can move flexibly on the construction site of the project, thereby reducing the requirements of the mobile sentinel robot on the project site.

[0039] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of the overall process of an autonomous navigation method for a mobile sentinel robot provided by the present invention.

[0041] Figure 2 A schematic diagram of the overall structure of an autonomous navigation system for a mobile sentinel robot provided by the present invention. DETAILED DESCRIPTION

[0042] like Figure 1-2 As shown, the present invention provides an autonomous navigation method for a mobile sentinel robot, such as Figure 1 As shown, including:

[0043] S1: Get the destination to be reached, set the navigation on the map, and move forward according to the navigation;

[0044] S2: Acquire the road surface information in the forward direction in real time, and determine whether the road surface information is passable by image processing. If passable, proceed according to the road surface information and repeat step S2. If not passable, obtain the portrait information in the road surface information and execute step S3.

[0045] S3: If the road information contains portrait information, a set broadcast voice is issued, and a reply voice is received. The reply voice is converted into text and then the meaning of the reply voice is extracted using a semantic extraction algorithm;

[0046] S4: Modify the original navigation according to the reply meaning to obtain a new navigation, and proceed according to the new navigation. If the destination is not reached, jump to step S2. If the destination is reached, end the operation.

[0047] The present invention obtains navigation according to a map, communicates with construction workers at the same time, recognizes the construction workers' voices to extract meanings therein, and corrects the navigation according to the meanings, and finally moves along the navigation, thereby improving the efficiency of reaching the destination. When the mobile sentinel robot encounters obstacles while walking on the construction site road surface, reasonable obstacle avoidance can be carried out through effective communication with the construction workers, so that the mobile sentinel robot can move flexibly on the construction site of the project, thereby reducing the requirements of the mobile sentinel robot on the project site.

[0048] In step S1, the destination to be reached is obtained. This can be done by inputting it or by the mobile sentinel robot itself calling it according to the progress of the project. Navigation is set on the map and the robot moves forward according to the navigation. These are routine operations of the map. Navigation is completed by the map software and the mobile sentinel robot moves forward according to the navigation instructions.

[0049] In step S2, the road surface information in the forward direction is obtained, including:

[0050] Get the current position of the mobile sentry robot;

[0051] Obtaining the position of the mobile sentry robot at the next time node according to the navigation;

[0052] Determine the forward direction according to the position of the mobile sentinel robot at the next time node;

[0053] Obtain road surface information based on the direction of travel.

[0054] The above-mentioned direction of travel is obtained through navigation, and then the road surface information in the direction of travel is obtained. In the present invention, the road surface information is obtained in the form of images, which are captured by the camera carried by the mobile sentinel robot itself.

[0055] In addition, in step S2, whether the road surface information is passable is determined by image processing, including:

[0056] Converting the road surface information into a road surface image, and extracting roads in the road surface image;

[0057] Determine whether there is an obstacle on the road in the road image, and determine the actual size and position of the obstacle;

[0058] Determine whether the road surface is passable based on the actual size and position of the obstacle.

[0059] The road surface information is converted into a road surface image by image processing, and then the road surface image is processed to extract obstacles and their sizes from the image.

[0060] In the present invention, if there is an obstacle, the mobile sentinel robot performs step S3, even if it starts to prepare to change the route to avoid the obstacle. If there is no obstacle, the mobile sentinel robot continues to move forward according to the original navigation. During the process of moving forward, it determines whether there is an obstacle in real time based on the road surface information.

[0061] The step of determining whether there is an obstacle on the road in the road surface image includes:

[0062] Extracting the color value of each pixel in the road surface image, and comparing each color value with the color value in a set range;

[0063] Filter out the pixel points corresponding to the color values ​​within the set range;

[0064] Extract pixels that are clustered together from the selected pixels through a clustering algorithm;

[0065] Determine whether the pixels gathered together vertically penetrate the road image. If so, there is no obstacle. If not, there is an obstacle.

[0066] Wherein, the determining of the actual size and position of the obstacle includes:

[0067] The obstacle contours in the road surface image are extracted by using a graphic contour extraction algorithm;

[0068] Obtain the size and position of obstacle outlines in the road image;

[0069] The actual size and position of the obstacle are calculated according to the focal length of the road surface image and the size and position of the obstacle in the road surface image.

[0070] The present invention extracts whether there is an obstacle on the road surface that can completely block the road surface by means of pixels, wherein the color value of the set range refers to the range composed of the values ​​of the color of the road surface in different exposure ranges. When these pixel points are gathered together, it means that a road surface is formed, and whether the pixel points gathered together vertically penetrate the road surface image can indicate whether there is an obstacle on the road surface that completely blocks the road surface. When there is an obstacle on the road surface that completely blocks the road surface, these pixel points gathered together cannot vertically penetrate the road surface image.

[0071] At the same time, the present invention determines the size and position of obstacles through the size and position of obstacles on the road surface, which needs to be calculated by means of focal length, and the focal length depends on the parameters in the camera of the mobile sentinel robot itself.

[0072] At the same time, in step S2, if the road is impassable, the portrait information in the road surface information is obtained, which paves the way for step S3 and people asking for directions. Only when there are people around can the voice asking for directions be effective.

[0073] In step S3, if the road surface information contains portrait information, it means that there are construction workers around the mobile sentinel robot. At this time, the mobile sentinel robot sends out a set broadcast voice, receives a reply voice, converts the reply voice into text, and uses a semantic extraction algorithm to extract the reply meaning of the reply voice.

[0074] The broadcast voice set in the present invention can be: Hello, can you tell me how to get to the destination? The restrictions of the answer can also be stated, such as please tell me in this way, turn left first, walk 50 meters and then turn right, etc.

[0075] In the present invention, the reply meaning includes equal reply times and reply directions, and each reply time and reply direction correspond to each other one by one.

[0076] In step 4, the original navigation is modified according to the reply meaning to obtain a new navigation, including:

[0077] The remaining navigation after the current position in the original navigation;

[0078] A new remaining navigation is set according to the response time and the response direction, and the correction of the original navigation is completed to obtain the new navigation.

[0079] The above method amends the original navigation according to the guidance of the construction personnel, and the subsequent navigation of the mobile sentinel robot is performed according to the new navigation. The mobile sentinel robot moves forward according to the new navigation, and when it does not reach the destination, it jumps to step S2. When it reaches the destination, the mobile sentinel robot ends the operation and has successfully reached the destination.

[0080] Based on the above-mentioned autonomous navigation method of a mobile sentinel robot, in order to implement the above-mentioned method through hardware foundation, the present invention also designs an autonomous navigation system of a mobile sentinel robot, including:

[0081] The navigation generation module is used to obtain the destination to be reached, set the navigation on the map, and proceed according to the navigation;

[0082] The road surface judgment module is used to obtain the road surface information of the forward direction in real time, and judge whether the road surface information is passable by means of image processing. If it is passable, the vehicle moves forward according to the road surface information and repeats the road surface judgment module. If it is not passable, the vehicle obtains the portrait information in the road surface information and executes the voice inquiry module.

[0083] The voice inquiry module is used to send out a set broadcast voice if there is portrait information in the road information, and receive a reply voice, and then use a semantic extraction algorithm to extract the reply meaning of the reply voice after converting the reply voice into text;

[0084] The navigation update module is used to correct the original navigation according to the reply meaning to obtain a new navigation, and move forward according to the new navigation. When the destination is not reached, it jumps to the road surface judgment module, and when the destination is reached, the operation ends.

[0085] The above modules cooperate with each other and work in coordination to jointly complete the above method, providing a hardware basis for the execution of the above-mentioned autonomous navigation method of a mobile sentinel robot.

[0086] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An autonomous navigation method for a mobile sentinel robot, characterized in that: include: S1: Get the destination to be reached, set the navigation on the map, and move forward according to the navigation; S2: Acquire the road surface information in the forward direction in real time, and determine whether the road surface information is passable by image processing. If passable, proceed according to the road surface information and repeat step S2. If not passable, obtain the portrait information in the road surface information and execute step S3. S3: If the road information contains portrait information, a set broadcast voice is issued, and a reply voice is received. The reply voice is converted into text and then the meaning of the reply voice is extracted using a semantic extraction algorithm; S4: Modify the original navigation according to the reply meaning to obtain a new navigation, and proceed according to the new navigation. If the destination is not reached, jump to step S2. If the destination is reached, end the operation.

2. The autonomous navigation method of a mobile sentinel robot according to claim 1, characterized in that: In step S2, the road surface information in the forward direction is obtained, including: Get the current position of the mobile sentry robot; Obtaining the position of the mobile sentry robot at the next time node according to the navigation; Determine the forward direction according to the position of the mobile sentinel robot at the next time node; Obtain road surface information based on the direction of travel.

3. The autonomous navigation method of a mobile sentinel robot according to claim 1, characterized in that: In step S2, whether the road surface information is passable is determined by image processing, including: Converting the road surface information into a road surface image, and extracting roads in the road surface image; Determine whether there is an obstacle on the road in the road image, and determine the actual size and position of the obstacle; Determine whether the road surface is passable based on the actual size and position of the obstacle.

4. The autonomous navigation method of a mobile sentinel robot according to claim 3, characterized in that: The determining whether there is an obstacle on the road in the road surface image includes: Extracting the color value of each pixel in the road surface image, and comparing each color value with the color value in a set range; Filter out the pixel points corresponding to the color values ​​within the set range; Extract pixels that are clustered together from the selected pixels through a clustering algorithm; Determine whether the pixels gathered together vertically penetrate the road image. If so, there is no obstacle. If not, there is an obstacle.

5. The autonomous navigation method of a mobile sentinel robot according to claim 3, characterized in that: The determining of the actual size and position of the obstacle includes: The obstacle contours in the road surface image are extracted by using a graphic contour extraction algorithm; Obtain the size and position of obstacle outlines in the road image; The actual size and position of the obstacle are calculated according to the focal length of the road surface image and the size and position of the obstacle in the road surface image.

6. The autonomous navigation method of a mobile sentinel robot according to claim 1, characterized in that: The reply meaning includes an equal number of reply times and reply directions, and each of the reply times and the reply directions corresponds to each other one by one.

7. The autonomous navigation method of a mobile sentinel robot according to claim 6, characterized in that: In step S4, the original navigation is modified according to the reply meaning to obtain a new navigation, including: The remaining navigation after the current position in the original navigation; A new remaining navigation is set according to the response time and the response direction, and the correction of the original navigation is completed to obtain the new navigation.

8. An autonomous navigation system for a mobile sentinel robot, characterized in that: include: The navigation generation module is used to obtain the destination to be reached, set the navigation on the map, and proceed according to the navigation; The road surface judgment module is used to obtain the road surface information of the forward direction in real time, and judge whether the road surface information is passable by means of image processing. If it is passable, the vehicle moves forward according to the road surface information and repeats the road surface judgment module. If it is not passable, the vehicle obtains the portrait information in the road surface information and executes the voice inquiry module. The voice inquiry module is used to send out a set broadcast voice if there is portrait information in the road information, and receive a reply voice, and then use a semantic extraction algorithm to extract the reply meaning of the reply voice after converting the reply voice into text; The navigation update module is used to correct the original navigation according to the reply meaning to obtain a new navigation, and move forward according to the new navigation. When the destination is not reached, it jumps to the road surface judgment module, and when the destination is reached, the operation ends.

Images