Method and device for determining region to be cleaned, dust cleaning apparatus, terminal for controlling cleaning robot, and storage medium

Patent metadata

Field	Value
Publication number	US-12035858-B2
Application number	US-202017288778-A
Country	US
Kind code	B2
Filing date	Jun 17, 2020
Priority date	Jun 18, 2019
Publication date	Jul 16, 2024
Grant date	Jul 16, 2024

Abstract

The present disclosure provides a method and a device for determining a region to be cleaned, a dust cleaning apparatus, a terminal for controlling a cleaning robot, and a storage medium. The method includes: determining, in an initial image of a photovoltaic panel, a gap region associated with a gap between cell pieces; assigning first preset values to pixel values of pixels located in the gap region of the initial image to obtain a first value-assigned image, wherein the first preset values are pixel values of pixels in the initial image that correspond to the cell pieces.

First claim

What is claimed is:

1. A method for determining a region to be cleaned, comprising:
determining a gap region belonging to a gap between solar cells in an initial image of a photovoltaic panel, wherein the photovoltaic panel comprises a plurality of solar cells;
assigning pixel values of pixels in the gap region of the initial image to a first preset value to obtain a first assignment image, wherein the first preset value is a pixel value of a pixel, corresponding the solar cells, in the initial image; and
determining a region to be cleaned with dust in the first assignment image, wherein determining the region to be cleaned with the dust in the first assignment image comprises:
according to the pixel values of the pixels in the first assignment image, determining feature values of the pixels in the first assignment image;
inputting the feature values into a preset classification model to determine whether the dust exists in the initial image; and
if the dust exists in the initial image, determining the region to be cleaned according to connected dust pixels corresponding to the dust.

2. The method according to claim 1 , wherein inputting the feature values into the preset classification model to determine whether the dust exists in the initial image, comprises:
dividing the first assignment image into a plurality of sub-regions;
determining a texture feature vector according to a histogram of feature values of pixels in each sub-region; and
inputting the texture feature vector into the preset classification model to determine whether the dust exists in the initial image.

3. The method according to claim 2 , wherein determining the region to be cleaned according to the connected dust pixels corresponding to the dust, comprises:
in pixels of the initial image, determining pixels, whose values of the texture feature vector are within a preset range, are the dust pixels corresponding to the dust;
assigning pixel values of the dust pixels to a second preset value to obtain a second assignment image; and
determining a connected region, comprising pixels with pixel values equal to the second preset value, in the second assignment image.

4. The method according to claim 3 , wherein determining the region to be cleaned according to the connected dust pixels corresponding to the dust, further comprises:
determining circumscribed rectangles of all connected regions as the region to be cleaned.

5. The method according to claim 4 , further comprising:
mapping the region to be cleaned onto the initial image, cutting out an image of the region to be cleaned, converting the image, which is cut out, into a grayscale image, determining a dust region based on the grayscale image, mapping the dust region onto the initial image, and outputting a located dust region.

6. The method according to claim 5 , wherein determining the dust region based on the grayscale image, comprises:
detecting the connected regions of the dust pixels, calculating circumscribed rectangles of the connected regions, and merging intersected circumscribed rectangles to obtain a set of final circumscribed rectangles, wherein the located dust region comprises regions, corresponding to all circumscribed rectangles in the set of final circumscribed rectangles, in the initial image.

7. The method according to claim 3 , wherein determining the region to be cleaned according to the connected dust pixels corresponding to the dust, further comprises:
determining a sub-circumscribed rectangle of each connected region; and
determining circumscribed rectangles of all sub-circumscribed rectangles as the region to be cleaned.

8. The method according to claim 2 , wherein the preset classification model comprises a support vector machine or a neural network.

9. The method according to claim 1 , wherein the feature values comprise a local binary pattern value.

10. The method according to claim 1 , wherein determining the gap region belonging to the gap between the solar cells in the initial image of the photovoltaic panel, comprises:
dividing the initial image into images of a plurality of color channels;
determining whether pixel values of pixels at a same position in images of respective color channels are greater than a third preset value;
determining that the pixels at the position in the initial image are gap pixels if the pixel values of the pixels at the same position in the images of respective color channels are greater than the third preset value; and
determining the gap region according to all gap pixels in the initial image.

11. A terminal for controlling a cleaning robot, configured to execute the method for determining the region to be cleaned according to claim 1 , inform the region to be cleaned, which is determined, to the cleaning robot, and instruct the cleaning robot to automatically clean the region to be cleaned.

12. The device according to claim 11 , wherein the processor is further configured to:
determine circumscribed rectangles of all connected regions as the region to be cleaned.

13. The device according to claim 11 , wherein the processor is further configured to:
determine a sub-circumscribed rectangle of each connected region; and
determine circumscribed rectangles of all sub-circumscribed rectangles as the region to be cleaned.

14. A device for determining a region to be cleaned, comprising a processor and a storage medium storing instructions, when the instructions are executed by the processor, the processor is configured to:
determine a gap region belonging to a gap between solar cells in an initial image of a photovoltaic panel, wherein the photovoltaic panel comprises a plurality of solar cells;
assign pixel values of pixels in the gap region of the initial image to a first preset value to obtain a first assignment image, wherein the first preset value is a pixel value of a pixel, corresponding the solar cells, in the initial image; and
determine a region to be cleaned with dust in the first assignment image, wherein when the processor is configured to determine the region to be cleaned with the dust in the first assignment image, the processor is further configured to:
determine feature values of the pixels in the first assignment image according to the pixel values of the pixels in the first assignment image;
input the feature values into a preset classification model to determine whether the dust exists in the initial image; and
determine the region to be cleaned according to connected dust pixels corresponding to the dust if the dust exists in the initial image.

15. A dust cleaning device, configured to automatically clean the region to be cleaned which is determined by the device for determining the region to be cleaned according to claim 14 .

16. The device according to claim 15 , wherein the processor is further configured to:
in pixels of the initial image, determine pixels, whose values of the texture feature vector are within a preset range, are the dust pixels corresponding to the dust;
assign pixel values of the dust pixels to a second preset value to obtain a second assignment image; and
determine a connected region, comprising pixels with pixel values equal to the second preset value, in the second assignment image.

17. The device according to claim 14 , wherein the processor is further configured to:
divide the first assignment image into a plurality of sub-regions;
determine a texture feature vector according to a histogram of feature values of pixels in each sub-region; and
input the texture feature vector into the preset classificati

Assignees

• Boe Technology Group Co Ltd

Inventors

• Xu Jingtao
• Suo Jianwen

Classifications

• Y02E10/50

  Photovoltaic [PV] energy · CPC title

• G06F18/2411

  based on the proximity to a decision surface, e.g. support vector machines · CPC title

• A47L2201/06

  Control of the cleaning action for autonomous devices; Automatic detection of the surface condition before, during or after cleaning · CPC title

• A47L9/2826

  the condition of the floor · CPC title

• G06V10/56

  relating to colour · CPC title