Environment model construction method applied to intelligent inspection robot

Abstract

The application discloses an environment model construction method applied to an intelligent inspection robot, relates to the technical field of robots, sets inspection key points in a target inspection area according to requirements, and deploys edge terminals at the inspection key points; infrared data at the inspection key points is acquired through the edge terminals, a local environment model corresponding to the edge terminals is constructed based on the acquired infrared data, an initial inspection position of the inspection robot is set, infrared data and image data of the initial inspection position are acquired through the inspection robot, a reference environment model with the inspection robot as the center is constructed according to the infrared data and the image data, and the local environment models corresponding to all the edge terminals are fitted with the reference environment model according to the inspection progress of the inspection robot, so that an inspection environment model is obtained; thereby, the inspection environment model construction process of the inspection robot is optimized, the process of generating the inspection environment model is smoother, and the efficiency of the inspection environment model construction is improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of robots, and particularly relates to an environment model construction method applied to an intelligent inspection robot. BACKGROUND

[0002] It is crucial for an intelligent inspection robot to construct an accurate environment model. In the early stage, a simple two-dimensional map construction method is applied, which can provide basic environment layout information but lacks three-dimensional information such as height and object shape, and cannot meet the requirements of complex scenes. With the development of technology, methods such as three-dimensional laser scanning appear, which can obtain three-dimensional point cloud data of the environment, but the processing process is complex, the data volume is large, and the real-time performance is poor.

[0003] How to reduce the burden of building a three-dimensional environment model on the inspection robot to improve the smoothness of the construction of the environment model is a problem to be solved, and therefore, the application provides an environment model construction method applied to an intelligent inspection robot. SUMMARY

[0004] The application aims to provide an environment model construction method applied to an intelligent inspection robot.

[0005] The application can be achieved by the following technical scheme: an environment model construction method applied to an intelligent inspection robot, comprising:

[0006] setting inspection key points in a target inspection area according to requirements, and deploying edge terminals at the inspection key points;

[0007] acquiring infrared data at the inspection key points through the edge terminals, and constructing a local environment model corresponding to the edge terminals based on the acquired infrared data;

[0008] setting an initial inspection position of the inspection robot, and acquiring infrared data and image data of the initial inspection position through the inspection robot;

[0009] constructing a reference environment model centered on the inspection robot according to the infrared data and the image data;

[0010] fitting the local environment models corresponding to the edge terminals to the reference environment model according to the inspection progress of the inspection robot to obtain an inspection environment model.

[0011] Further, the process of setting inspection key points in a target inspection area according to requirements and deploying edge terminals at the inspection key points comprises:

[0012] The target inspection area is provided with a robot inspection route for the inspection robot to perform inspection;

[0013] According to actual conditions, the inspection key points are set on the robot inspection route.

[0014] The edge terminal is arranged at the inspection key point, and each edge terminal is provided with a communication coverage range and a data detection range.

[0015] Further, the process of acquiring infrared data at the inspection key point through the edge terminal and constructing a local environment model corresponding to the edge terminal based on the acquired infrared data comprises:

[0016] The infrared data at the inspection key point is periodically acquired through the edge terminal arranged at the inspection key point;

[0017] The data detection range of the edge terminal is acquired, and the overlapping part of the data detection range and the robot inspection route is marked as a target area corresponding to the edge terminal;

[0018] The position of the edge terminal is recorded as a reference point, and a three-dimensional space coordinate system with the reference point as the origin is constructed;

[0019] According to the obtained infrared data, the obstacle information in the target area is detected, and the coordinate position of the detected obstacle in the three-dimensional space coordinate system is marked;

[0020] The coordinate position of each detected obstacle is compared with the coordinate position of the obstacle obtained in the previous period, if there is an obstacle with the same coordinate position, the corresponding obstacle is marked as a fixed obstacle, otherwise, it is marked as a non-fixed obstacle;

[0021] According to the obstacle information of the fixed obstacle, a corresponding obstacle model is generated at the corresponding position in the three-dimensional space coordinate system, thereby obtaining a local environment model corresponding to the target area.

[0022] Further, the initial position of the inspection robot is set, and the process of acquiring infrared data and image data at the initial position of the inspection robot through the inspection robot comprises:

[0023] An initial inspection position for the inspection robot to perform inspection is set;

[0024] After the inspection robot is placed at the initial inspection position, the inspection robot acquires infrared data and image data in the corresponding direction of the initial inspection position according to the robot inspection route.

[0025] Further, the process of constructing a reference environment model centered on the inspection robot according to the infrared data and the image data comprises:

[0026] A three-dimensional space coordinate system centered on the inspection robot is constructed;

[0027] The image data obtained by the inspection robot is converted into image frames, the image frames are sorted according to time, and the image frames are rasterized, and the rasterized image data is converted into a grayscale image;

[0028] The obtained grayscale image is subjected to feature extraction, and obstacles and obstacle types contained in the image data are identified, the obstacle types including non-fixed obstacles and fixed obstacles;

[0029] According to the contours of the identified fixed obstacles and non-fixed obstacles, corresponding obstacle models are generated, and the obstacle models are mapped into a three-dimensional coordinate system;

[0030] Further, according to the infrared data, the obstacle information in the infrared scanning range of the inspection robot is detected, and the coordinate positions of the detected obstacles are marked in the three-dimensional coordinate system;

[0031] The coordinate positions of the obstacles detected by the infrared data are fitted with the positions of the obstacles identified by the image data to obtain a corresponding reference environment model.

[0032] Further, fitting the coordinate positions of the obstacles detected by the infrared data with the positions of the obstacles identified by the image data means that when there is a difference between the coordinates of the obstacles identified by the two, the intermediate value of the two obstacle coordinates is taken as the new obstacle coordinate.

[0033] Further, the process of identifying the obstacles and obstacle types contained in the image data includes:

[0034] Features in each image frame are extracted, and the contours of the obstacles in each image frame are marked according to the extracted features;

[0035] The contours of each obstacle extracted in adjacent image frames are matched, and the positions of the obstacles with the same contour are compared;

[0036] If the positions of the obstacles with the same contour change, it indicates that the corresponding obstacle is a non-fixed obstacle, and if the positions of the obstacles with the same contour do not change, it indicates that the corresponding obstacle is a fixed obstacle.

[0037] Further, according to the inspection progress of the inspection robot, the local environment model corresponding to each edge terminal is fitted with the reference environment model to obtain an inspection environment model, the process including:

[0038] When the inspection robot enters the communication coverage range of the edge terminal, the local environment model corresponding to the edge terminal is called;

[0039] According to the current position of the inspection robot and the position of the edge terminal, a coordinate deviation is obtained;

[0040] According to the coordinate deviation, the coordinates of the obstacle model in the local environment model are updated, and according to the updated coordinates, the obstacle model is mapped to the corresponding coordinates in the reference environment model, so that the fitting of the local environment model and the reference environment model is completed, and the corresponding inspection environment model is obtained.

[0041] Compared with the prior art, the beneficial effects of the present application are:

[0042] By setting an edge terminal on the robot inspection route, the infrared data in the data detection range is periodically acquired through the edge terminal, so that the distribution of the fixed obstacles in the data detection range is identified, so that when the inspection robot enters the communication coverage range of the edge terminal, the model of the fixed obstacles in the edge terminal can be directly sent to the model of the inspection robot, thereby optimizing the inspection environment model construction process of the inspection robot, making the process of generating the inspection environment model more smooth, and improving the efficiency of the inspection environment model construction. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0044] Figure 1 Flowchart of the present application. DETAILED DESCRIPTION

[0045] As Figure 1 shown, the environment model construction method applied to the intelligent inspection robot comprises:

[0046] Setting inspection key points in the target inspection area according to requirements, and deploying edge terminals at the inspection key points;

[0047] Acquiring infrared data at the inspection key points through the edge terminals, and constructing a local environment model corresponding to the edge terminals based on the acquired infrared data;

[0048] Setting an initial inspection position of the inspection robot, and acquiring infrared data and image data of the initial inspection position through the inspection robot;

[0049] Constructing a reference environment model centered on the inspection robot according to the infrared data and the image data;

[0050] According to the inspection progress of the inspection robot, fitting the local environment model corresponding to each edge terminal with the reference environment model to obtain an inspection environment model.

[0051] It needs to be further explained that, in the specific implementation process, the process of setting the inspection key points in the target inspection area according to the needs and deploying the edge terminal at the inspection key points includes:

[0052] The target inspection area is provided with a robot inspection route for the inspection robot to perform inspection;

[0053] According to the actual situation, the inspection key points are set on the robot inspection route; it needs to be noted that the inspection key points are provided with ports for the inspection robot to obtain inspection data;

[0054] The edge terminal is set at the inspection key point, and the edge terminal can be in communication connection with the inspection robot, so as to complete the data interaction between the edge terminal and the inspection robot, and each edge terminal is provided with a communication coverage range and a data detection range.

[0055] It needs to be further explained that, in the specific implementation process, the process of obtaining the infrared data at the inspection key point through the edge terminal, and constructing the local environment model corresponding to the edge terminal based on the obtained infrared data includes:

[0056] The infrared data at the inspection key point is periodically obtained through the edge terminal set at each inspection key point; it needs to be noted that the periodic acquisition of infrared data means that a group of infrared data is obtained every certain period of time;

[0057] According to the order of the edge terminal on the robot inspection route, each edge terminal is labeled, denoted as i, where i=1, 2, …n, and n is an integer greater than 0;

[0058] The data detection range of the edge terminal with the label i is obtained, and the overlapping part of the data detection range and the robot inspection route is marked as a target area corresponding to the edge terminal;

[0059] The position of the edge terminal is recorded as a reference point, and a three-dimensional space coordinate system with the reference point as the origin is constructed;

[0060] According to the obtained infrared data, the obstacle information in the target area is detected, and the coordinate position of the detected obstacle is marked in the three-dimensional space coordinate system;

[0061] The coordinate positions of the detected obstacles are compared with the coordinate positions of the obstacles obtained in the previous period, if there is an obstacle with the same coordinate position, the corresponding obstacle is marked as a fixed obstacle, otherwise, it is marked as a non-fixed obstacle;

[0062] According to the obstacle information of the fixed obstacle, a corresponding obstacle model is generated at a corresponding position in a three-dimensional space coordinate system, so as to obtain a local environment model corresponding to the target region.

[0063] It needs to be further explained that, in the specific implementation process, the process of setting the initial inspection position of the inspection robot and obtaining the infrared data and image data of the initial inspection position by the inspection robot includes:

[0064] An initial inspection position for the inspection robot to perform inspection is set; it needs to be noted that the initial inspection position is located on the robot inspection route;

[0065] After the inspection robot is placed at the initial inspection position, the inspection robot obtains the infrared data and image data of the direction corresponding to the initial inspection position according to the robot inspection route.

[0066] It needs to be further explained that, in the specific implementation process, the process of constructing a reference environment model centered on the inspection robot according to the infrared data and image data includes:

[0067] A three-dimensional space coordinate system centered on the inspection robot is constructed;

[0068] The image data obtained by the inspection robot is converted into image frames, the image frames are sorted according to time, the image frames are rasterized, and the rasterized image data is converted into a grayscale image;

[0069] The obtained grayscale image is feature-extracted to identify the obstacles and the obstacle types contained in the image data, the obstacle types including fixed obstacles and non-fixed obstacles; it needs to be noted that, in the present application, an edge detection algorithm is used to identify the obstacles in the image data, and a Sobel operator is used as the edge detection operator; the process of identifying the obstacles in the image data by using the edge detection algorithm is a common technical means for those skilled in the art, and will not be described here;

[0070] According to the contours of the identified fixed obstacles and non-fixed obstacles, corresponding obstacle models are generated, and the obstacle models are mapped to the three-dimensional space coordinate system;

[0071] According to the infrared data, the obstacle information in the infrared scanning range of the inspection robot is detected, and the coordinate positions of the detected obstacles are marked in the three-dimensional space coordinate system;

[0072] The coordinate position of the obstacle detected by the infrared data is fitted with the position of the obstacle identified by the image data to obtain a corresponding reference environment model; wherein, when the coordinates of the obstacles identified by the two are different, the middle value of the two obstacle coordinates is taken as the new obstacle coordinate;

[0073] It should be noted that the specific process of identifying the obstacles and obstacle types contained in the image data includes:

[0074] According to the order of the image frames, the image frames are sequentially labeled, denoted as j, where j=1, 2, …, m, and m is an integer greater than 0;

[0075] Features are extracted from each image frame, and the contours of the obstacles in each image frame are labeled according to the extracted features;

[0076] The contours of each obstacle extracted from the image frame labeled j=1 are matched with the contours of each obstacle extracted from the image frame labeled j=2, and the positions of the obstacles with the same contour are compared;

[0077] If the positions of the obstacles with the same contour change, it indicates that the corresponding obstacle is a non-fixed obstacle, and if the positions of the obstacles with the same contour do not change, it indicates that the corresponding obstacle is a fixed obstacle;

[0078] The contours of each obstacle extracted from the image frame labeled j=3 are matched with the contours of each obstacle extracted from the image frame labeled j=2, and the positions of the obstacles with the same contour are compared, and so on.

[0079] It should be further noted that in the specific implementation process, according to the inspection progress of the inspection robot, the local environment model corresponding to each edge terminal is fitted with the reference environment model to obtain the inspection environment model, which includes:

[0080] When the inspection robot enters the communication coverage range of the edge terminal, the local environment model corresponding to the edge terminal is called;

[0081] According to the current position of the inspection robot and the position of the edge terminal, the coordinate deviation is obtained;

[0082] The coordinates of the obstacle model in the local environment model are updated according to the coordinate deviation, and the obstacle model is mapped to the corresponding coordinates in the reference environment model according to the updated coordinates, thereby completing the fitting of the local environment model and the reference environment model, and obtaining the corresponding inspection environment model.

[0083] It should be noted that when the inspection robot enters the communication coverage range of the edge terminal, the inspection robot will no longer generate the corresponding obstacle model for the fixed obstacles in the identified obstacles, but only generate the obstacle model of the non-fixed obstacles, and directly call the obstacle model in the local environment model corresponding to the edge terminal for the fixed obstacles, so that the inspection robot can reduce a part of the model generation process after entering the communication coverage range of the edge terminal, thereby realizing node optimization of the inspection environment model construction process, and making the process of generating the inspection environment model more smooth.

[0084] It should be noted that in actual situations, there may be a communication blind area on the robot inspection route, i.e., the corresponding area is not within the communication coverage range of any edge terminal. When the inspection robot is in the communication blind area, the generation of the obstacle model of the fixed obstacle is resumed, and the obstacle model generation process at the initial inspection position is referred to, which is not described herein.

[0085] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes, without departing from the scope of the technical solution of the present application. Any modification or equivalent replacement of the above embodiments according to the technical essence of the present application still belongs to the scope of the technical solution of the present application.

Claims

1.A method for constructing an environment model applied to a smart inspection robot, characterized in that, The method comprises the following steps: setting inspection key points in a target inspection area according to requirements, and deploying edge terminals at the inspection key points; acquiring infrared data at the inspection key points through the edge terminals, and constructing a local environment model corresponding to the edge terminals based on the acquired infrared data; setting an initial inspection position of an inspection robot, and acquiring infrared data and image data of the initial inspection position through the inspection robot; constructing a benchmark environment model centered on the inspection robot according to the infrared data and the image data; fitting the local environment model corresponding to each edge terminal with the benchmark environment model according to the inspection progress of the inspection robot, to obtain an inspection environment model. 2.The method for constructing an environment model applied to a smart inspection robot according to claim 1, wherein, The process of setting inspection key points in a target inspection area according to requirements, and deploying edge terminals at the inspection key points comprises: a robot inspection route for an inspection robot to perform inspection is set in the target inspection area; inspection key points are set on the robot inspection route according to actual conditions; edge terminals are set at the inspection key points, and each edge terminal is provided with a communication coverage range and a data detection range. 3.The method for constructing an environment model applied to a smart inspection robot according to claim 2, characterized in that, The process of acquiring infrared data at the inspection key points through the edge terminals, and constructing a local environment model corresponding to the edge terminals based on the acquired infrared data comprises: periodically acquiring infrared data at the inspection key points through the edge terminals set at the inspection key points; acquiring the data detection range of the edge terminal, and marking the overlapping part of the data detection range and the robot inspection route as a target area corresponding to the edge terminal; marking the position of the edge terminal as a reference point, and constructing a three-dimensional space coordinate system with the reference point as the origin; detecting obstacle information in the target area according to the acquired infrared data, and marking the coordinate position of the detected obstacle in the three-dimensional space coordinate system; comparing the coordinate position of each detected obstacle with the coordinate position of the obstacle obtained in the previous period, if there is an obstacle with the same coordinate position, the corresponding obstacle is marked as a fixed obstacle, otherwise, it is marked as a non-fixed obstacle; generating a corresponding obstacle model at the corresponding position in the three-dimensional space coordinate system according to the obstacle information of the fixed obstacle, thereby obtaining a local environment model corresponding to the target area. 4.The method for constructing an environment model applied to a smart inspection robot according to claim 3, characterized in that, The process of setting an initial inspection position of an inspection robot, and acquiring infrared data and image data of the initial inspection position through the inspection robot comprises: setting an initial inspection position for the inspection robot to perform inspection; after placing the inspection robot at the initial inspection position, the inspection robot acquires infrared data and image data in the corresponding direction of the initial inspection position according to the robot inspection route. 5.The method for constructing an environment model applied to a smart inspection robot according to claim 4, characterized in that, The process of constructing a benchmark environment model centered on the inspection robot according to the infrared data and the image data comprises: constructing a three-dimensional space coordinate system centered on the inspection robot; converting the image data obtained by the inspection robot into image frames, sorting the image frames according to time, and performing rasterization processing on the image frames, and converting the rasterized image data into a grayscale image; The obtained grayscale image is subjected to feature extraction, and obstacles and obstacle types contained in the image data are identified, the obstacle types including non-fixed obstacles and fixed obstacles; According to the contours of the identified fixed obstacles and non-fixed obstacles, corresponding obstacle models are generated, and the obstacle models are mapped into a three-dimensional spatial coordinate system; According to the infrared data, obstacle information in the infrared scanning range of the inspection robot is detected, and the coordinate positions of the detected obstacles are marked in the three-dimensional spatial coordinate system; The coordinate positions of the obstacles detected by the infrared data are fitted with the positions of the obstacles identified by the image data, and a corresponding reference environment model is obtained. 6.The method for constructing an environment model applied to a smart inspection robot according to claim 5, wherein, Fitting the coordinate positions of the obstacles detected by the infrared data with the positions of the obstacles identified by the image data means that when there is a difference between the coordinates of the obstacles identified by the two, the intermediate value of the two obstacle coordinates is taken as the new obstacle coordinate. 7.The method for constructing an environment model applied to a smart inspection robot according to claim 5, wherein, The process of identifying obstacles and obstacle types contained in the image data includes: Features in each image frame are extracted, and the contours of obstacles in each image frame are marked according to the extracted features; The contours of each obstacle extracted in adjacent image frames are matched, and the positions of obstacles with the same contour are compared; If the positions of obstacles with the same contour change, it indicates that the corresponding obstacle is a non-fixed obstacle, and if the positions of obstacles with the same contour do not change, it indicates that the corresponding obstacle is a fixed obstacle. 8.The method for constructing an environment model applied to a smart inspection robot according to claim 7, wherein, According to the inspection progress of the inspection robot, the local environment model corresponding to each edge terminal is fitted with the reference environment model to obtain an inspection environment model, including: When the inspection robot enters the communication coverage range of the edge terminal, the local environment model corresponding to the edge terminal is called; According to the current position of the inspection robot and the position of the edge terminal, a coordinate deviation is obtained; According to the coordinate deviation, the coordinates of the obstacle models in the local environment model are updated, and according to the updated coordinates, the obstacle models are mapped into corresponding coordinates in the reference environment model, thereby completing the fitting of the local environment model and the reference environment model, and obtaining a corresponding inspection environment model.

Images