CN111242939B - Method for identifying state of pressing plate - Google Patents

Abstract

The invention provides a method for identifying a pressing plate state, which comprises the following steps: s channels in an HSV color space of the input graph are taken out, and the threshold 80 and the maximum 255 are binarized; morphological operation is carried out by using morph l ogyEx function, the kernel selects a rectangle with the size of 9x9, and the image is subjected to open operation; searching the Contours by using f i ndContours functions, and storing all the found Contours Contours; taking a minimum circumscribed rectangle BoundRects for each element of the contents; calculating the area median of all BoundREcts, and carrying out first screening on the deviation domain of each element limiting BoundREcts and BoundREcts to screen TarARcts 1; performing second screening on the limited aspect ratio of each element of TarRNA 1 to screen TarRNA 2; judging the switching state of the pressing plate according to the aspect ratio of each element of TarRacts 2; the x-coordinate of each element of TarRacts 2 is sequenced in an increasing mode, adjacent element differences are calculated, and the adjacent differences are converted into percentage changes dIff; the y-coordinate of each element of Tarrcts 2 is incrementally ordered, adjacent element differences are calculated, and the adjacent differences are converted into percent changes dIff.

Description

Method for identifying state of pressing plate

Technical Field

The invention relates to the technical field of power system inspection assistance, in particular to a method for identifying a pressing plate state.

Background

In 110kv substations, there are usually 10-20 protection cabinets, the number of the protection cabinets of 220kv and 500kv substations can reach hundreds, the number of the pressure plate switches of each protection cabinet is up to fifty-four, and inspection work needs to be performed on the pressure plate switches. However, the existing manual inspection process has the problems of large workload, poor timeliness, paperiness of work records and the like, and inspection personnel can have errors due to physical and psychological quality, responsibility, technical level, working experience and other factors, so that potential safety hazards are left, and even serious safety disasters are caused. Due to the development of the existing machine vision technology and network technology, a clamp plate part of a protection screen cabinet can be photographed through a camera, and then an image is processed through a processor to identify the on-off state of the clamp plate. However, the prior art does not have a method that can accurately recognize the state of the platen.

Disclosure of Invention

The invention provides a method for identifying the state of a pressing plate, which can effectively solve the problems.

The invention is realized in the following way:

the method for identifying the state of the pressing plate is characterized by comprising the following steps of:

s331, taking out an S channel in an HSV color space of the input diagram, and carrying out binarization of a threshold value 80 and a maximum value 255;

s332, performing morphological operation by using a morphyoyEx function, selecting a rectangle with the size of 9x9 by a kernel, and performing operation on an image;

s333, searching the contour by utilizing a findContours function, and storing all the found contour contents;

s334, taking a minimum circumscribed rectangle BoundRects for each element of the contents;

s335, calculating the area median of all BoundREcts, and carrying out first screening on each element of BoundREcts to limit the deviation domain of BoundREcts and BoundREcts, so as to screen TarRacts 1;

s336, performing second screening on the limited aspect ratio of each element of TarRNA 1 to screen TarRNA 2;

s337, judging the switching state of the pressing plate according to the aspect ratio for each element of TarRacts 2;

s338, carrying out incremental sequencing on the x coordinate of each element of TarRacts 2, calculating the difference of adjacent elements, and converting the adjacent differences into percentage change diff;

s339, the y coordinates of each element of TarRacts 2 are incrementally ordered, adjacent element differences are calculated, and the adjacent differences are converted into percentage change diff.

The beneficial effects of the invention are as follows: the method for identifying the state of the pressing plate can accurately identify the state of the pressing plate, and the accuracy of identification can reach more than 99.9.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some examples of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 to 23 are process images of background filtering, rotation correction, and platen state recognition processing in the platen state automatic checking method according to the embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments.

Referring to fig. 1 to 23, an embodiment of the present invention provides a method for automatically checking a state of a platen, including the steps of:

s1, scanning a two-dimensional code corresponding to a protection screen cabinet through a handheld scanning device, and acquiring the setting state of a pressure plate switch on the protection screen cabinet from a server;

s2, acquiring a photo of the protection screen cabinet through the handheld scanning equipment and uploading the photo to a server;

s3, the server sequentially carries out background filtering, rotation correction and pressing plate state identification on the photo, and transmits an identification result back to the handheld scanning device;

s4, the handheld scanning equipment compares the identification result with the set state and displays the comparison result on the handheld scanning equipment;

s5, the inspection personnel re-inspect according to the comparison result, and upload the result to the server for storage after confirming that the result is correct.

Referring to fig. 1, in step S3, the step of performing background filtering on the photo includes:

s311, an S channel in an HSV color space of an input photo is acquired, a threshold value 80 and a maximum value 255 are binarized, and a picture is obtained as shown in FIG. 2:

s312, morphological operation is carried out by using a morphyoyEx function, a kernel selects a rectangle with the size of 9x9, and an image is subjected to open operation (corrosion is carried out before expansion), so that a picture is shown in FIG. 3:

and (3) carrying out an opening operation:

img_open＝open(img_THR，element)＝dnate(erode(img_THR elemet))

firstly, corroding:

re-expanding:

s313, searching the Contours by utilizing the findContours function, and storing all the found Contours contents to obtain a picture as shown in fig. 4:

Contour8＝findContours(img_open)。

s314, taking the minimum circumscribed rectangle BoundRects for each element of the contents, and obtaining a picture as shown in FIG. 5:

BoundRects＝{boundRect|boundRect

＝Rect(min(contour.x)，min(contour，y)，max(contour，x)-min(contour.x)，max(contour，y)-min(contour，y))，contour∈Contoura)。

s315, calculating the area median of all BoundREcts (without area boundary selection, so that the algorithm is applicable to distribution boards with various pixel sizes), and performing first screening on each element of BoundREcts to define the deviation domain of BoundREcts and the deviation domain thereof, and screening out TarRacts 1, as shown in FIG. 6:

the area size set after incremental sequencing according to the area size of BoundRects is terms

terma(BoundRects)＝Ascending order(BoundRects，aree)

Calculating median of terms

First screening

S316, performing a second screening on the limited aspect ratio of each element of Tarrcts 1 to screen Tarrcts 2, as shown in FIG. 7;

TarRects2＝{rect|2*rect.height＞rect.width，rect∈TarRects1)。

s317, calculating and normalizing a histogram H for each element of Tarrcts 2 by using a calcHist function, comparing the histogram with a previously prepared red and Huang Yaban template by using a compactreHist function, defining an upper score limit, filtering non-red and Huang Yaban, and performing third screening of Tarrcts 3, as shown in FIG. 9:

the comparison histogram:

s318, calculating four corner points for each element of TarRact 3, merging and storing into a point container;

s319, taking the minimum circumscribed rectangle filter_rect of all points in the point container, adjusting the boundary range, and filling the part outside the area in the input photo with white to obtain a photo with a filtered background, as shown in FIG. 10.

As a further improvement, in step S3, the step of performing rotation correction on the photograph includes:

s321, filtering the photo of the background to obtain a minimum area rectangle (minAreate), and then obtaining a two-dimensional box (CvBox 2D) to obtain a single rotation angle, as shown in FIG. 11; solving for

Average rotation angle avg_angle.

Contour of pressing plate after three times of screening under the same index of background filtering:

Contoura2＝{contour|contour＝Contgura[rect.index]，rect∈TarRect3)

average rotation angle:

s322, selecting the larger number in the longer and wider width of the input diagramSetting the square with the length and the width of canvas as canvas, and filling black into the gap between the input diagram and the canvas by using a copyMakeBorder function, as shown in FIG. 12;

square canvas side length:

length and width of upper, lower, left and right filling areas:

dx＝(canvas_length-img_BGFliter.cols)/2

dy＝(canvas_length-img_BGFilter rowa)/2

thus, the filled canvas:

canvaa＝Rect(0，0，canvaa_length，canvaa_length)。

s323, rotating the input diagram by using the warp Affine function, wherein the rotation angle is avg_angle, and the canvas center is denoted as center, as shown in FIG. 13.

The original coordinates (x, y) are transformed into new coordinates (x ', y'), the intermediate transformation matrix being denoted M.

The basic formula of the transformation of the warp affine,

translation transformation, if

x′＝x+t _x

Y′＝y+t _y

Then

Rotation conversion, if the rotation angle is

In the image processing of opencv, all processing is performed from the origin, the origin of the image defaults to the upper left corner of the image, and when we perform rotation processing on the image, the midpoint of the image is generally used as the axis, so the following processing is needed: the axle center is moved to the origin to make rotation transformation, and finally the upper left corner is set as the origin of the image.

arc2(x′，y′)＝M3ximg.BGFnter(x，y)。

S324, calculating the size of the minimum circumscribed rectangle img_Recor of the rotated image, and cutting off redundant frames of the canvas, as shown in FIG. 14:

output img_record length and width:

img_Recor.width＝src2.cola*aba(cos(avg_angle))+arc2，rowa*aba(sin(avg_angle))img_Recor，height＝src2.cda*aba(ain(avg_angle))+src2，rowa*aba(coa(avg_angle))

cutting canvas area length and width:

dx2＝(canvas_length-img_Recor，width)/2

dy2＝(canvas_length-img__Recor.height)/2

thus, img_record is output:

img_Recor＝canvas(dx2，dy2，img_Recor.width，img_Recor.height)。

as a further improvement, in step S3, the step of performing platen state recognition on the photo includes:

s331 to S336 are the same as steps S311 to S316, and refer to fig. 15 to 20.

S337, judging the switching state of the pressing plate according to the aspect ratio for each element of TarRacts 2;

please refer to fig. 21:

opening:

TarRecta_on＝{rect|2*rect.width＞rect.height，rect∈TarRects2)

and (3) closing:

TarRects_off＝{rect|2*rect，width≤rect，heightrect∈TarRecta2}。

s338, carrying out incremental sequencing on the x coordinate of each element of TarRacts 2, calculating the difference of adjacent elements, and converting the adjacent differences into percentage change diff; s339, the y-coordinate increment of each element of tarrces 2 is sequenced, the adjacent element differences are calculated, and the adjacent differences are converted into percentage change diff, please refer to fig. 22 and 23:

setting a threshold value diffVal, and entering the next cluster when diff is larger than diffVal.

TarRects2＝Ascending order(TarRecta2)

The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, and various modifications and variations may be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (1)

1. The method for identifying the state of the pressing plate is characterized by comprising the following steps of:

s331, taking out an S channel in an HSV color space of the input diagram, and carrying out binarization of a threshold value 80 and a maximum value 255;

s332, performing morphological operation by using a morphyoyEx function, selecting a rectangle with the size of 9x9 by a kernel, and performing operation on an image;

s333, searching the contour by utilizing a findContours function, and storing all the found contour contents;

s334, taking a minimum circumscribed rectangle BoundRects for each element of the contents;

s335, calculating the area median of all BoundREcts, and carrying out first screening on each element of BoundREcts to limit the deviation domain of BoundREcts and BoundREcts, so as to screen TarRacts 1;

s336, performing second screening on the limited aspect ratio of each element of TarRNA 1 to screen TarRNA 2;

s337, judging the switching state of the pressing plate according to the aspect ratio for each element of TarRacts 2;

s338, carrying out incremental sequencing on the x coordinate of each element of TarRacts 2, calculating the difference of adjacent elements, and converting the adjacent differences into percentage change diff;

s339, the y coordinates of each element of TarRacts 2 are incrementally ordered, adjacent element differences are calculated, and the adjacent differences are converted into percentage change diff.