Search

CN-121982279-A - Photovoltaic panel characteristic detection method and device, electronic equipment and storage medium

CN121982279ACN 121982279 ACN121982279 ACN 121982279ACN-121982279-A

Abstract

The invention discloses a photovoltaic panel characteristic detection method, a photovoltaic panel characteristic detection device, electronic equipment and a storage medium. The method comprises the steps of obtaining a photovoltaic panel image acquired by a cleaning device at the current moment based on a binocular depth camera carried on the cleaning device, cleaning the photovoltaic panel according to a preset route by the cleaning device, carrying out feature recognition on a target object in the photovoltaic panel image to obtain parameter information of the target object, wherein the parameter information comprises an object type, a target area and a first confidence coefficient, matching a feature detection algorithm aiming at the target object based on the object type of the target object, processing the feature information of the target area based on the feature detection algorithm, determining a second confidence coefficient of the target object, determining the target confidence coefficient based on the first confidence coefficient and the second confidence coefficient, and determining the accuracy of the target object of the target area based on the target confidence coefficient. The method and the device remarkably improve the effectiveness of feature extraction and the overall detection accuracy.

Inventors

  • WANG FAN

Assignees

  • 星逻智能科技(苏州)有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (13)

  1. 1. A method for detecting characteristics of a photovoltaic panel, the method comprising: acquiring a photovoltaic panel image acquired by a cleaning device at the current moment based on a binocular depth camera carried on the cleaning device, wherein the cleaning device is used for cleaning the photovoltaic panel according to a preset route; Performing feature recognition on a target object in the photovoltaic panel image to obtain parameter information of the target object, wherein the parameter information comprises an object type, a target area and a first confidence coefficient; matching a feature detection algorithm for the target object based on the object type of the target object, processing feature information of the target area based on the feature detection algorithm, and determining a second confidence coefficient of the target object; Determining a target confidence based on the first confidence and the second confidence, and determining the accuracy of a target object of the target region based on the target confidence, wherein the accuracy is used for representing the probability that the target object is true.
  2. 2. The method according to claim 1, wherein the target object comprises a target edge and/or a photovoltaic panel slit, the target edge being an edge of a photovoltaic panel to which the cleaning apparatus is being directed; correspondingly, performing feature recognition on the target object in the photovoltaic panel image to obtain parameter information of the target object, wherein the method comprises the following steps: performing feature recognition on a target object of the photovoltaic panel image based on an object recognition model to obtain object feature data and feature confidence corresponding to the target object; If the feature confidence coefficient is larger than a first preset confidence coefficient, marking a data frame on the object feature data, determining a target area of the target object, and determining the feature confidence coefficient larger than the first preset confidence coefficient as the first confidence coefficient; and determining the object type of the target object based on the object characteristic data.
  3. 3. The method of claim 2, wherein if the target object is a target edge, the feature information of the target region includes depth information, wherein the processing the feature information of the target region based on the feature detection algorithm, determining the second confidence level of the target object, comprises: Determining a central area of the target area according to preset feature extraction conditions, wherein the preset feature extraction conditions are used for describing rule information for extracting the central area; Determining edge pixel points of the target edge based on the depth information of the central area; And determining a second confidence of the target object based on the number of edge pixel points and the total pixel point number of the central area.
  4. 4. A method according to claim 3, wherein said determining edge pixels of said target edge based on depth information of said central region comprises: performing gradient calculation on the depth information of the central region, and determining a plurality of gradient amplitudes of the central region; if the gradient amplitude is larger than the preset amplitude, determining a central area corresponding to the gradient amplitude larger than the preset amplitude as a depth abrupt change area; And determining the pixel points of each depth mutation area as edge pixel points of the target edge.
  5. 5. The method of claim 4, wherein after determining a center region corresponding to the gradient magnitude greater than a preset magnitude as a depth abrupt region, the method further comprises: and carrying out morphological treatment on each depth mutation region to obtain updated depth mutation regions.
  6. 6. The method of claim 2, wherein if the target object is a photovoltaic panel slit, the processing the feature information of the target area based on the feature detection algorithm to determine a second confidence level of the target object comprises: Expanding the target area according to a preset expansion ratio to obtain an expansion area; and determining a second confidence of the target object based on the height information of the extension area.
  7. 7. The method of claim 6, wherein determining the second confidence level of the target object based on the height information of the extended region comprises: determining the second confidence level using the formula : ; Wherein, the The height information is the average height of the gaps of the photovoltaic panel, h is the average height of the gaps of the photovoltaic panel, and alpha is the high tolerance coefficient.
  8. 8. The method of claim 2, wherein if the target object is a target edge, the photovoltaic panel image includes depth information, the depth information being a depth value corresponding to each pixel, and wherein before the processing the feature information of the target area based on the feature detection algorithm, determining the second confidence level of the target object, the method further comprises: Determining an area with a preset height above the bottom edge of the photovoltaic panel image as a core area, and dividing the core area into a preset number of grid areas; Determining pixel points larger than a first depth threshold value in the depth values of the pixel points of the grid area as first pixel points, and determining a first pixel point occupation ratio of the number of the first pixel points and the total pixel points of the grid area; If the first pixel point duty ratio is larger than or equal to a target preset proportion, determining the grid area to be in a first state, wherein the first state is used for describing that the grid area does not have characteristic information of a target edge; if the first pixel point duty ratio is smaller than a target preset proportion, determining the pixel points larger than a second depth threshold value in the depth values of the pixel points of the grid area as second pixel points, and determining the second pixel point duty ratio of the number of the second pixel points to the total pixel points of the grid area, wherein the first depth threshold value is smaller than the second depth threshold value; Determining state information of the grid region based on the second pixel point duty ratio and the average depth value of the grid region, wherein the state information is used for describing the characteristic duty ratio condition of characteristic information of a target edge in the grid region; and determining whether to process the characteristic information of the target area based on the characteristic detection algorithm based on the state information of all the grid areas, and determining a second confidence of the target object.
  9. 9. The method of claim 8, wherein determining the status information of the grid region based on the second pixel point duty cycle and the average depth value of the grid region comprises: Determining a third confidence coefficient of the grid region based on the second pixel point duty ratio, wherein the third confidence coefficient is used for reflecting the characteristic duty ratio condition of the characteristic information of the target edge in the grid region; If the third confidence coefficient is larger than a preset confidence coefficient threshold value and the average depth value is larger than or equal to a third depth threshold value, determining the state information of the grid region to be a second state, wherein the second state is used for describing that the feature ratio of the feature information of the target edge in the grid region is a first feature ratio; If the average depth value is smaller than a third depth threshold value, determining the state information of the grid region to be a third state, wherein the third state is used for describing that the feature ratio of the feature information of the target edge in the grid region is a second feature ratio; and if the third confidence coefficient is larger than a preset confidence coefficient threshold value and the average depth value is larger than or equal to a fourth depth threshold value, determining the state information of the grid region to be a first state, wherein the fourth depth threshold value is larger than the third depth threshold value.
  10. 10. The method of claim 9, wherein determining whether to process the feature information of the target region based on the feature detection algorithm based on the state information of all the grid regions, determining the second confidence level of the target object comprises: And if the number of the second states is greater than or equal to the preset number, determining to process the feature information of the target area based on the feature detection algorithm, and determining a second confidence coefficient of the target object.
  11. 11. A photovoltaic panel feature detection apparatus, the apparatus comprising: the system comprises an image acquisition module, a cleaning device and a control module, wherein the image acquisition module is used for acquiring a photovoltaic panel image acquired by the cleaning device at the current moment based on a binocular depth camera carried on the cleaning device; The parameter determining module is used for carrying out feature recognition on a target object in the photovoltaic panel image to obtain parameter information of the target object, wherein the parameter information comprises an object type, a target area and a first confidence coefficient; The feature processing module is used for matching a feature detection algorithm aiming at the target object based on the object type of the target object, processing feature information of the target area based on the feature detection algorithm and determining a second confidence coefficient of the target object; The analysis module is used for determining a target confidence coefficient based on the first confidence coefficient and the second confidence coefficient, determining the accuracy of a target object of the target area based on the target confidence coefficient, and the accuracy is used for representing the probability that the target object is true.
  12. 12. An electronic device, the electronic device comprising: at least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the photovoltaic panel feature detection method of any one of claims 1-10.
  13. 13. A computer readable storage medium storing computer instructions for causing a processor to perform the photovoltaic panel feature detection method of any one of claims 1-10.

Description

Photovoltaic panel characteristic detection method and device, electronic equipment and storage medium Technical Field The present invention relates to the field of image processing technologies, and in particular, to a method and apparatus for detecting characteristics of a photovoltaic panel, an electronic device, and a storage medium. Background The photovoltaic panel is assembled by utilizing a plurality of photovoltaic modules, so that the generated energy can be greatly improved. With the continuous expansion of the scale of photovoltaic generators and the diversification of deployment areas, the problem of efficiency loss caused by dust accumulation on the surface of photovoltaic panels is increasingly prominent. Thus, photovoltaic panel cleaning is necessary to improve photovoltaic system energy efficiency. At present, a full-automatic cleaning device is generally adopted to clean a photovoltaic panel, and in the process of cleaning the photovoltaic panel by the full-automatic cleaning device, the edge of the photovoltaic panel and a gap between the panels can influence the route of the full-automatic cleaning device for executing a cleaning task, so that the full-automatic cleaning device needs to accurately identify the characteristics so as to ensure that the cleaning device can timely adjust the travelling direction and the operation gesture. At present, the characteristic recognition of the photovoltaic panel generally carries out the characteristic recognition of the target through the collected image to determine the edge and the gap between the photovoltaic panels on the photovoltaic panel, but the characteristic recognition of the target has the defects of limited precision and frequent false detection. Disclosure of Invention The invention provides a photovoltaic panel feature detection method, a photovoltaic panel feature detection device, electronic equipment and a storage medium, so that the effectiveness of feature extraction and the overall detection accuracy are remarkably improved. According to an aspect of the present invention, there is provided a photovoltaic panel feature detection method, the method comprising: acquiring a photovoltaic panel image acquired by a cleaning device at the current moment based on a binocular depth camera carried on the cleaning device, wherein the cleaning device is used for cleaning the photovoltaic panel according to a preset route; Performing feature recognition on a target object in the photovoltaic panel image to obtain parameter information of the target object, wherein the parameter information comprises an object type, a target area and a first confidence coefficient; matching a feature detection algorithm for the target object based on the object type of the target object, processing feature information of the target area based on the feature detection algorithm, and determining a second confidence coefficient of the target object; And determining a target confidence based on the first confidence and the second confidence, and determining the accuracy of a target object of the target region based on the target confidence, wherein the accuracy is used for representing the probability that the target object is true. According to another aspect of the present invention, there is provided a photovoltaic panel feature detection apparatus, the apparatus comprising: the system comprises an image acquisition module, a cleaning device and a control module, wherein the image acquisition module is used for acquiring a photovoltaic panel image acquired by the cleaning device at the current moment based on a binocular depth camera carried on the cleaning device; The parameter determining module is used for carrying out feature recognition on a target object in the photovoltaic panel image to obtain parameter information of the target object, wherein the parameter information comprises an object type, a target area and a first confidence coefficient; The feature processing module is used for matching a feature detection algorithm aiming at the target object based on the object type of the target object, processing feature information of the target area based on the feature detection algorithm and determining a second confidence coefficient of the target object; The analysis module is used for determining a target confidence coefficient based on the first confidence coefficient and the second confidence coefficient, determining the accuracy of a target object of the target area based on the target confidence coefficient, and the accuracy is used for representing the probability that the target object is true. According to another aspect of the present invention, there is provided an electronic apparatus including: at least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the photovoltaic pa