A cleaning robot positioning method and a cleaning robot using the same

Abstract

The present invention relates to a positioning method for a cleaning robot, which is characterized by comprising the following steps: Step 1, obtaining a priori map of a room ceiling; Step 2, during the actual cleaning process, the cleaning robot obtains the position of the current position in the priori map of the room ceiling through the image of the room ceiling photographed at the current position: the image of the room ceiling photographed at the current position is called a real-time image, the intersection of two diagonals in the real-time image is used as a real-time position point, the boundary of the room ceiling in the real-time image is obtained, the distance from the real-time position point to four straight lines in the boundary of the room ceiling in the real-time image is calculated, and then according to the priori map of the room ceiling, the specific position of the real-time position point in the priori map of the entire room ceiling is obtained. Compared with the prior art, the advantages of the present invention are: it can improve the positioning accuracy of the cleaning robot and reduce the error caused by calculating the positioning point.

Description

Technical Field

[0001] The invention relates to a cleaning robot positioning method and a cleaning robot adopting the method. Background Art

[0002] In order to achieve a higher coverage rate and complete the path planning and navigation tasks more efficiently during the cleaning process, the cleaning robot needs to have a higher positioning accuracy. However, due to the large number of obstacles in the room, there are certain errors in positioning and mapping. Many cleaning robots position the robot by installing a camera in the center to shoot the ceiling vertically. For example, the Chinese patent "Robot positioning and map construction system based on ceiling image information" with patent number ZL201410149081.1 and the Chinese patent application "Visual positioning method of indoor mobile robot based on ceiling" with application number CN201610725330.6 both set a camera that can take pictures of the ceiling on the cleaning robot, and continuously compare the characteristic straight lines and characteristic points in the two ceiling photos, and convert the comparison results into the motion rotation angle and motion displacement increment of the mobile robot in the world coordinate system, so as to achieve the self-positioning of the mobile robot. By taking the ceiling as a reference target, compared with other objects, the ceiling is not easily blocked and the elements in the photo are relatively monotonous, which is convenient for the image processing unit to extract contours and analyze photos.

[0003] However, there are few special marks on the ceiling. When using the above method for feature extraction and matching, matching errors often occur due to similar features, affecting the positioning accuracy. In addition, by comparing the feature points of two frames of images to calculate its own position, there are certain errors in the feature extraction, matching and calculation processes. As the cleaning robot moves, the positioning error will continue to accumulate, resulting in increased navigation and mapping errors. Therefore, the existing technology needs to be further improved. Summary of the invention

[0004] The primary technical problem to be solved by the present invention is to provide a positioning method for a cleaning robot with small error and precise positioning in view of the above-mentioned prior art.

[0005] A further technical problem to be solved by the present invention is to provide a cleaning robot with small error and precise positioning.

[0006] The technical solution adopted by the present invention to solve the above-mentioned primary technical problem is: a positioning method of a cleaning robot, characterized by comprising the following steps:

[0007] Step 1: Get the room ceiling prior map. The specific method is:

[0008] The camera installed on the cleaning robot is used to take the room ceiling image, and the points in the room ceiling image whose grayscale values ​​are obviously different from the surrounding grayscale values ​​are extracted as identification points. The extracted identification points are fitted into four straight lines using a fitting method, so as to obtain the boundary of the entire room ceiling, and then obtain the room ceiling prior map;

[0009] Step 2: During the actual cleaning process, the cleaning robot obtains the position of the current position in the prior map of the room ceiling through the image of the room ceiling captured at the current position, which specifically includes:

[0010] Step 2-1, the room ceiling image captured at the current position is called a real-time image, the intersection of two diagonals in the real-time image is used as the position of the cleaning robot in the real-time image at this time, and the position point is marked as a real-time position point;

[0011] Step 2-2, extracting points in the real-time image where the grayscale value of the ceiling is significantly different from the grayscale value of the surrounding areas as identification points, fitting these extracted identification points into straight lines using a fitting method, and matching the fitted straight lines with the boundaries of the prior map of the room ceiling, thereby obtaining the boundaries of the room ceiling in the real-time image;

[0012] Step 2-3, calculate the distance from the real-time position point to the four straight lines in the room ceiling boundary in the real-time image, and then obtain the specific position of the real-time position point in the entire room ceiling prior map based on the room ceiling prior map.

[0013] As an improvement, in step 1, the cleaning robot adopts an autonomous exploration method to move within a certain range in the room to be cleaned, obtains feature information of the entire ceiling in the entire room, and thus obtains a priori map of the room ceiling.

[0014] Further improvement, in step 1, the extraction method used when extracting the identification points in the room ceiling image includes but is not limited to the FAST feature detection method, the SURF feature detection method, the SIFT feature extraction method, and the line feature extraction method.

[0015] Further improvement, in the step 1, after obtaining the prior map of the room ceiling, the length and width information of the entire room are obtained.

[0016] The technical solution adopted by the present invention to solve the above-mentioned further technical problems is: a cleaning robot includes a body, a camera is provided on the top of the body, and a walking mechanism is provided on the body, characterized in that: a controller electrically connected to the walking mechanism and the camera is provided in the body, and the controller adopts the above-mentioned positioning method to position the cleaning robot.

[0017] Compared with the prior art, the advantages of the present invention are:

[0018] 1. It can improve the positioning accuracy of the cleaning robot and reduce the error caused by calculating the positioning point;

[0019] 2. It can reduce the cumulative positioning error generated after cleaning for a period of time;

[0020] 3. It can reduce the error caused by image matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of a positioning method for a cleaning robot in an embodiment of the present invention.

[0022] Figure 2 A diagram of a method for obtaining a real-time location point for a real-time image taken by a cleaning robot.

[0023] Figure 3 Schematic diagram of fitting boundary lines to a real-time image taken by a cleaning robot.

[0024] Figure 4 Schematic diagram of fitting boundary lines to form the ceiling boundary in a real-time image taken by the cleaning robot.

[0025] Figure 5 Schematic diagram of a method for calculating the distance between a real-time position point and a ceiling boundary in a real-time image taken by a cleaning robot. DETAILED DESCRIPTION

[0026] The present invention is described in detail below with reference to the accompanying drawings.

[0027] like Figure 1 As shown, the present invention provides a positioning method for a cleaning robot, which is characterized by comprising the following steps:

[0028] Step 1: Get the room ceiling prior map. The specific method is:

[0029] The cleaning robot adopts an autonomous exploration method to move within a certain range in the room to be cleaned, and uses a wide-angle camera or a 360° fisheye camera arranged on the top of the cleaning robot to capture the room ceiling image, and adopts a FAST feature detection method or a SURF feature detection method or a SIFT feature extraction method or a line feature extraction method to extract points in the room ceiling image whose grayscale values ​​are obviously different from those of the surrounding areas as identification points, and then adopts a fitting method to fit these extracted identification points into four straight lines, so as to obtain the boundary of the entire room ceiling, and then obtain a priori map of the room ceiling, and obtain the length and width information of the four boundaries in the room ceiling;

[0030] Step 2: During the actual cleaning process, the cleaning robot takes an image of the room ceiling at the current position, and then obtains the position of the current position in the prior map of the room ceiling, including:

[0031] Step 2-1: The room ceiling image captured at the current position is called a real-time image. The intersection of the two diagonals in the real-time image is used as the position of the cleaning robot in the real-time image at this time, and the position point is marked as a real-time position point. Take a real-time image captured by the cleaning robot as an example, and use the intersection of the diagonals as the real-time position point. Figure 2 As shown;

[0032] Step 2-2, using the FAST feature detection method or the SURF feature detection method or the SIFT feature extraction method or the line feature extraction method, extract points in the real-time image where the grayscale value of the ceiling is significantly different from the grayscale value of the surrounding areas as identification points, and use the fitting method to fit these extracted identification points into straight lines; also taking a real-time image taken by the cleaning robot as an example, the fitted straight line is shown in Figure 3 As shown; if the fitted straight line cannot form a complete rectangle or square, a complete rectangle or square can be obtained by drawing a virtual extension line based on the existing straight line, see Figure 4 As shown, according to the length and width information of the room, the fitted boundary line is matched with the length information of the room in the prior map of the room ceiling, so as to obtain the boundary of the room ceiling in the real-time image;

[0033] Step 2-3, finally calculate the distance from the real-time position point to the four straight lines in the room ceiling boundary in the real-time image, see Figure 5 As shown, based on the room ceiling prior map, the specific position of the real-time position point in the entire room ceiling prior map is obtained.

[0034] The present invention also provides a cleaning robot, including a body, a camera is provided on the top of the body, a walking mechanism is provided on the body, and a controller electrically connected to the walking mechanism and the camera is provided in the body. The controller uses the above-mentioned positioning method to position the cleaning robot.

Claims

1. A positioning method for a cleaning robot, Features The steps include: Step 1: Get the room ceiling prior map. The specific method is: The camera installed on the cleaning robot is used to take the room ceiling image, and the points in the room ceiling image whose grayscale values ​​are obviously different from the surrounding grayscale values ​​are extracted as identification points. The extracted identification points are fitted into four straight lines using a fitting method, so as to obtain the boundary of the entire room ceiling, and then obtain the room ceiling prior map; Step 2: During the actual cleaning process, the cleaning robot obtains the position of the current position in the prior map of the room ceiling through the image of the room ceiling captured at the current position, which specifically includes: Step 2-1, the room ceiling image captured at the current position is called a real-time image, the intersection of two diagonals in the real-time image is used as the position of the cleaning robot in the real-time image at this time, and the position point is marked as a real-time position point; Step 2-2, extracting points in the real-time image where the grayscale value of the ceiling is significantly different from the grayscale value of the surrounding areas as identification points, fitting these extracted identification points into straight lines using a fitting method, and matching the fitted straight lines with the boundaries of the prior map of the room ceiling, thereby obtaining the boundaries of the room ceiling in the real-time image; Step 2-3, calculate the distance from the real-time position point to the four straight lines in the room ceiling boundary in the real-time image, and then obtain the specific position of the real-time position point in the entire room ceiling prior map based on the room ceiling prior map.

2. The cleaning robot positioning method according to claim 1, Features: In the step 1, the cleaning robot moves within a certain range in the room to be cleaned in an autonomous exploration manner to obtain feature information of the entire ceiling in the room, thereby obtaining a priori map of the room ceiling.

3. The cleaning robot positioning method according to claim 1, Features: In the step 1, the extraction methods used when extracting the identification points in the room ceiling image include a FAST feature detection method, a SURF feature detection method, a SIFT feature extraction method, and a line feature extraction method.

4. The cleaning robot positioning method according to claim 1, Features: In the step 1, after obtaining the prior map of the room ceiling, the length and width information of the entire room is obtained; in the step 2-2, according to the length and width information of the room, the fitted straight line is matched with the length information of the room in the prior map of the room ceiling, so as to obtain the boundary of the room ceiling in the real-time image.

5. A cleaning robot, comprising a body, a camera is arranged on the top of the body, and a walking mechanism is arranged on the body. Features: A controller electrically connected to the walking mechanism and the camera is disposed in the body, and the controller uses the positioning method as claimed in claim 1 to position the cleaning robot.

Images