CN106057700A - Method for detecting edge red film of solar cell panel - Google Patents

Abstract

The invention discloses a method for detecting edge red film of a solar cell panel by using a machine vision detection system. The method comprises the following steps: acquiring the RGB image of a solar cell panel through the CCD camera of a machine vision detection system; generating a cell panel region through a back light source and extracting cell panel gate lines through an R channel; combining the two steps to remove external and white interference; and finally converting for an HSV model. According to the invention, an edge tonal variance based detection method is provided to calculate the color variance of the solar cell panel. When the color variance of the solar cell panel at the edge region represents the value determined as the disclosed figure, the solar cell panel is determined to be an edge red film. The method of the invention uses a machine vision system, adopts a detection method based on the image mean value and the image variance to select the edge red film of a solar cell panel, which replaces the traditional manual detection method, effectively improves the appearance quality of a cell panel and reduces labor intensity of manual work and the manual fragmentation rate, as well as the manual pollution to the cell, therefore, improving productivity.

Description

A kind of detection method on red of the limit of solar battery sheet

Technical field

The present invention relates to the detection method on red of the limit of a kind of solar battery sheet, belong to field of photovoltaic technology.

Background technology

At present, the whole world faces serious fossil energy crisis and environmental crisis, and solar energy is clean as one, pollution-free and Abundant natural energy resources, is widely used.But, the production technology of solar battery sheet is more complicated, causes producing The cell piece come is inevitably generated colour deficient problem, as red in limit, ash sheet etc..Inspection for cell piece colour deficient Surveying, major part uses the mode of manual detection, personnel's fatiguability after time length, and each personnel's touchstone is inconsistent, meeting Cause cell piece to mix shelves, cell piece can be polluted simultaneously.In a practical situation, the appearance color defect of cell piece, not only can affect Photovoltaic module presentation quality and service life, and photoelectric transformation efficiency can be reduced.

Summary of the invention

In order to solve the problems referred to above, the invention provides the detection method on red of the limit of a kind of solar battery sheet, this Bright technology contents is as follows:

The detection method on red of the limit of a kind of solar battery sheet, utilizes Machine Vision Inspecting System, and its step is as follows:

(1), solar battery sheet RGB image is gathered by the ccd video camera of Machine Vision Inspecting System；

(2), cell piece region is produced by backlight；

(3), cell piece grid line is extracted by R passage；

(4), step (2) and step (3) combine and remove peripheral and White lnterfere；

(5) HSV model, finally it is converted into, containing three vector values in described HSV model, i.e. tone H, saturation S, bright Degree V；

(6), use detection method based on edge tone variance identification, calculate the variance of the tone H of solar battery sheet, As the variance ＞ 8 of solar battery sheet marginal area tone H, it is determined that this sheet cell piece edge color is uneven, for red of limit.

Above-mentioned steps (4) is removed peripheral and White lnterfere method as follows:

Collect solar battery sheet RGB image by step (1), the cell piece grid line that step (3) is extracted, use based on Periphery and White lnterfere are carried out segmentation and remove by the dividing method of threshold values, owing to amount of image information is relatively big, therefore from the figure collected Directly obtain target area in Xiang, and set peripheral and White lnterfere threshold values T value, and by T value, image is divided into more than T pixel Group with less than T pixel group, threshold values input picture F (i, j) to output image G (x, y) such as down conversion:

Thus the periphery in image and White lnterfere are split away, and get rid of its interference to detection.

The method being converted into HSV model in above-mentioned steps (5) is as follows:

Tone H that R, G, B value of solar battery sheet RGB image is converted in hsv color model by equation below, full With degree S, brightness V:

$\left\{\begin{array}{c}max=max(R,G,B);\\ min=min(R,G,B);\\ V=max(R,G,B)/255;\\ S=(\max -\min )/max;\\ if(R=max)H=(G-B)/(max-min)\&times;60;\\ if(G=\max )H=120+(B-R)/(max-min)\&times;60;\\ if(B=\max )H=240+(R-G)/(\max -min)\&times;60;\\ f(H<0)H=H+360;\end{array}\right.$

In hsv color model, it is assumed that have N number of pixel, each pixel in one piece of solar battery sheet solid images Tone H be defined as H (N_n), n=1,2,3 ... N, the Mean Value Formulas of tone this component of H is:

$\stackrel{\&OverBar;}{H}=\frac{H\left(N_1\right)+H\left(N_2\right)+...+H\left(N_n\right)}{n}.$

The formula of the variance calculating tone H in above-mentioned steps (6) is as follows:

${\mathrm{\&sigma;}}_H^2=\frac{{\left(H\right(N_1)-\stackrel{\&OverBar;}{H})}^2+{\left(H\right(N_2)-\stackrel{\&OverBar;}{H})}^2+...+{\left(H\right(N_n)-\stackrel{\&OverBar;}{H})}^2}{n}.$

The beneficial effect that the present invention is reached:

The present invention utilizes Vision Builder for Automated Inspection, uses detection method based on image average and variance to solar battery sheet Red of limit sort, replace Traditional Man detection mode, be effectively improved cell piece presentation quality, reduce manual work strong Degree, reduces manual operation fragment rate, reduces the artificial pollution to cell piece, improves productivity ratio.

Detailed description of the invention

Below in conjunction with embodiment, the invention will be further described.Following example are only used for clearly illustrating this Bright technical scheme, and can not limit the scope of the invention with this.

The detection method on red of the limit of a kind of solar battery sheet, utilizes Machine Vision Inspecting System, and its step is as follows:

(1), solar battery sheet RGB image is gathered by the ccd video camera of Machine Vision Inspecting System；

(2), cell piece region is produced by backlight；

(3), cell piece grid line is extracted by R passage；

(4), step (2) and step (3) combine and remove peripheral and White lnterfere；

(5) HSV model, finally it is converted into, containing three vector values in described HSV model, i.e. tone H, saturation S, bright Degree V；

(6), use detection method based on edge tone variance identification, calculate the variance of the tone H of solar battery sheet, As the variance ＞ 8 of solar battery sheet marginal area tone H, it is determined that this sheet cell piece edge color is uneven, for red of limit.

Above-mentioned steps (4) is removed peripheral and White lnterfere method as follows:

Collect solar battery sheet RGB image by step (1), the cell piece grid line that step (3) is extracted, use based on Periphery and White lnterfere are carried out segmentation and remove by the dividing method of threshold values, owing to amount of image information is relatively big, therefore from the figure collected Directly obtain target area in Xiang, and set peripheral and White lnterfere threshold values T value, and by T value, image is divided into more than T pixel Group with less than T pixel group, threshold values input picture F (i, j) to output image G (x, y) such as down conversion:

Thus the periphery in image and White lnterfere are split away, and get rid of its interference to detection.

The method being converted into HSV model in above-mentioned steps (5) is as follows:

Tone H that R, G, B value of solar battery sheet RGB image is converted in hsv color model by equation below, full With degree S, brightness V:

$\left\{\begin{array}{c}max=max(R,G,B);\\ min=min(R,G,B);\\ V=max(R,G,B)/255;\\ S=(\max -\min )/max;\\ if(R=max)H=(G-B)/(max-min)\&times;60;\\ if(G=\max )H=120+(B-R)/(max-min)\&times;60;\\ if(B=\max )H=240+(R-G)/(\max -min)\&times;60;\\ f(H<0)H=H+360;\end{array}\right.$

In hsv color model, it is assumed that have N number of pixel, each pixel in one piece of solar battery sheet solid images Tone H be defined as H (N_n), n=1,2,3 ... N, the Mean Value Formulas of tone this component of H is:

$\stackrel{\&OverBar;}{H}=\frac{H\left(N_1\right)+H\left(N_2\right)+...+H\left(N_n\right)}{n}.$

The formula of the variance calculating tone H in above-mentioned steps (6) is as follows:

${\mathrm{\&sigma;}}_H^2=\frac{{\left(H\right(N_1)-\stackrel{\&OverBar;}{H})}^2+{\left(H\right(N_2)-\stackrel{\&OverBar;}{H})}^2+...+{\left(H\right(N_n)-\stackrel{\&OverBar;}{H})}^2}{n}.$

The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art For Yuan, on the premise of without departing from the technology of the present invention principle, it is also possible to make some improvement and deformation, these improve and deformation Also should be regarded as protection scope of the present invention.

Claims (4)

1. the detection method on red of the limit of a solar battery sheet, it is characterised in that utilize Machine Vision Inspecting System, its step Rapid as follows:

(1), solar battery sheet RGB image is gathered by the ccd video camera of Machine Vision Inspecting System；

(2), cell piece region is produced by backlight；

(3), cell piece grid line is extracted by R passage；

(4), step (2) and step (3) combine and remove peripheral and White lnterfere；

(5) HSV model, finally it is converted into, containing three vector values in described HSV model, i.e. tone H, saturation S, brightness V；

(6), use detection method based on edge tone variance identification, calculate the variance of the tone H of solar battery sheet, when too Sun can the variance ＞ 8 of cell piece marginal area tone H time, it is determined that this sheet cell piece edge color is uneven, for red of limit.

The detection method on red of the limit of a kind of solar battery sheet the most according to claim 1, it is characterised in that: described step (4) remove peripheral and White lnterfere method suddenly as follows:

Collect solar battery sheet RGB image, the cell piece grid line that step (3) is extracted by step (1), use based on threshold values Dividing method periphery and White lnterfere carried out segmentation remove, owing to amount of image information is relatively big, therefore from the image collected Directly obtain target area, and set peripheral and White lnterfere threshold values T value, and image is divided into more than T pixel group by T value and Less than T pixel group, threshold values input picture F (i, j) to output image G (x, y) such as down conversion:

Thus the periphery in image and White lnterfere are split away, and get rid of its interference to detection.

The detection method on red of the limit of a kind of solar battery sheet the most according to claim 1, it is characterised in that: described step Suddenly the method being converted into HSV model in (5) is as follows:

Tone H that R, G, B value of solar battery sheet RGB image is converted in hsv color model by equation below, saturation S, brightness V:

$\left\{\begin{array}{c}max=max(R,G,B);\\ min=min(R,G,B);\\ V=max(R,G,B)/255;\\ S=(\max -\min )/max;\\ if(R=max)H=(G-B)/(max-min)\&times;60;\\ if(G=\max )H=120+(B-R)/(max-min)\&times;60;\\ if(B=\max )H=240+(R-G)/(\max -min)\&times;60;\\ f(H<0)H=H+360;\end{array}\right.$

In hsv color model, it is assumed that have N number of pixel, the color of each pixel in one piece of solar battery sheet solid images H is adjusted to be defined as H (N_n), n=1,2,3 ... N, the Mean Value Formulas of tone this component of H is:

$\stackrel{\&OverBar;}{H}=\frac{H\left(N_1\right)+H\left(N_2\right)+...+H\left(N_n\right)}{n}.$

The detection method on red of the limit of a kind of solar battery sheet the most according to claim 1, it is characterised in that: described step The formula of the variance suddenly calculating tone H in (6) is as follows:

${\mathrm{\&sigma;}}_H^2=\frac{{\left(H\right(N_1)-\stackrel{\&OverBar;}{H})}^2+{\left(H\right(N_2)-\stackrel{\&OverBar;}{H})}^2+...+{\left(H\right(N_n)-\stackrel{\&OverBar;}{H})}^2}{n}.$