CN112408153A - Method for monitoring movement amount of steel wire rope in elevator braking process - Google Patents
Method for monitoring movement amount of steel wire rope in elevator braking process Download PDFInfo
- Publication number
- CN112408153A CN112408153A CN202011076684.5A CN202011076684A CN112408153A CN 112408153 A CN112408153 A CN 112408153A CN 202011076684 A CN202011076684 A CN 202011076684A CN 112408153 A CN112408153 A CN 112408153A
- Authority
- CN
- China
- Prior art keywords
- wire rope
- steel wire
- detection
- straight line
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B7/00—Other common features of elevators
- B66B7/12—Checking, lubricating, or cleaning means for ropes, cables or guides
- B66B7/1207—Checking means
- B66B7/1215—Checking means specially adapted for ropes or cables
- B66B7/1223—Checking means specially adapted for ropes or cables by analysing electric variables
Abstract
The invention provides a method for monitoring the movement amount of a steel wire rope in the braking process of an elevator, which is characterized by comprising the following steps of: the industrial camera is used for shooting the traction steel wire rope group in the braking process, the shot image is transmitted into the computer, the computer calculates the movement amount of the traction steel wire rope by using an image recognition algorithm, and the result is displayed on the display equipment. The invention realizes non-contact measurement, the measuring device does not influence the performance of the brake, the traction braking device is not required to be modified, the measured object can be found out according to the image, the requirement on the installation precision is reduced, the influence of environment change on a detection system is also reduced, and the task of monitoring the movement of the steel wire rope for a long time can be completed.
Description
Technical Field
The invention relates to monitoring of the movement amount of a steel wire rope of an elevator traction braking device, and belongs to the technical field of elevator detection.
Background
When the elevator is in use, the steel wire rope wound on the traction sheave continuously rubs with the traction sheave groove. For an old elevator traction sheave, the sheave groove of the old elevator traction sheave is gradually abraded, the friction between the traction sheave and a steel wire rope is reduced, and the traction capacity is reduced. When the traction capacity is reduced to a certain degree, the traction wheel and the steel wire rope slip, the elevator car cannot be normally driven, and if the traction wheel slips in the operation process, safety accidents can occur. In addition, the brake shoe rubs with the brake wheel when braking each time, so that the brake shoe is worn, the braking capability is reduced, the brake wheel slips, and safety accidents also occur.
The performance of the traction braking device of an elevator is measured by the braking distance. The braking distance is the amount of movement of the car during braking. The car is directly connected with the steel wire rope, and the braking distance is equal to the movement amount of the steel wire rope in the braking process, so that the car is a result of the comprehensive action of an elevator braking system and an elevator traction system. The amount of movement of the brake sheave relative to the brake shoe plus the amount of movement of the steel rope relative to the traction sheave during braking of the elevator equals the total amount of movement of the steel rope (the brake sheave and the traction sheave can be regarded as a rigid connection). The amount of movement of the brake wheel relative to the brake shoe represents the braking capacity of the elevator. The amount of movement of the wire rope relative to the traction sheave represents the traction capacity. When the braking capability or the traction capability of the elevator is reduced, the movement amount of the steel wire rope in the braking process is increased. Therefore, the movement amount of the steel wire rope is detected, and the braking capability and the traction capability of the elevator can be comprehensively reflected.
For measuring the movement amount of the steel wire rope, the conventional detection device is commonly used in the following ways: and a small wheel with large friction is used for clamping the steel wire rope, and an angle encoder is installed on the small wheel. When the steel wire rope moves, the small wheel is driven to rotate, and the angle encoder counts. And calculating the moving distance of the steel wire rope according to the rotation angle and the diameter of the small wheel. The method is low in cost, easy to implement and suitable for field detection. But the small wheel and the steel wire rope also slide, and the measurement accuracy is lower and lower due to loose installation and abrasion of the small wheel as the service life is prolonged. Therefore, this method is not suitable for long-term monitoring of the amount of movement of the wire rope.
For example: the utility model discloses an application number is CN201821457677.8, the utility model discloses a (announcement number) is CN208916530U discloses an elevator braking distance detection device, it relates to elevator detection technical field, concretely discloses an elevator braking distance detection device, including the data sender, make and stop signal wiring, first distance detection sensor and second distance detection sensor, the data sender respectively with make and stop signal wiring, first distance sensor and second distance sensor are connected, be used for sending detection data to the detection personnel, make and stop signal wiring be used for connecting the brake relay of elevator in order to detect in order to receive the braking signal of elevator, first distance detection sensor is equipped with the detection gyro wheel, the detection gyro wheel is used for cooperating the displacement distance with the wire rope that tows wire rope with the overspeed governor of elevator, second distance detection sensor is equipped with detection magnet, the center that detection magnet is used for adsorbing at the wheel of elevator detects the slip distance of wire rope in wheel department And (5) separating.
The invention patent with the application number of CN201210163507.X and the publication number of CN102674101B discloses a displacement measuring device for an elevator steel wire rope, wherein an optical image displacement detection system is formed by applying an optical mouse chip component, a guide wheel with a rope groove is arranged on an instrument frame, a component of the guide wheel is made of a magnetic material, and the guide wheel has magnetic attraction to a measured steel wire rope, so that the measuring device and the measured steel wire rope keep a specified mutual position, and the non-contact displacement measurement is realized. The device is suitable for measuring the running speed and the stroke of the elevator, is simple and convenient to operate, and has higher detection precision.
The invention patent application with the application number of CN201410238968.8 and the publication number of CN104016201A discloses an elevator car absolute position detection device, which comprises a traction wheel, a guide wheel, a traction steel wire rope, a car, a counterweight, an elevator main control board, a light projector, a light receiver and a position controller, wherein the light projector and the light receiver are arranged on two sides of the traction steel wire rope, the position controller is communicated with the elevator main control board, the light receiver comprises a photosensitive element and an image recognition element, the light projector projects light beams to the traction steel wire rope, images are projected on the photosensitive element of the light receiver when the traction steel wire rope moves, the image recognition element of the light receiver periodically compares the image and samples the characteristic, and generates a corresponding displacement signal according to the change of the sampling characteristic, the position controller carries out direction judgment and calculation on the displacement signal, and then detects the position of the elevator car. The elevator car absolute position detection device is simple and practical, high in precision and free of the problem of detection error caused by slipping.
The invention patent application with the application number of CN201910653566.7 and the publication number of CN110240038A discloses a device and a method for detecting the slippage of an elevator traction sheave. The device comprises an elevator traction system, an image acquisition device, a wireless transmission device and a monitoring center; an elevator traction system comprising: the traction sheave, the traction sheave motor, a steel wire rope with scale marks and a control cabinet; the traction sheave motor is electrically connected with the traction sheave; the steel wire rope is placed in the rope groove of the traction wheel and is matched with the rope groove of the traction wheel; the control cabinet is electrically connected with the traction sheave motor and is used for reading the revolution of the traction sheave motor; the image acquisition device is arranged above the traction sheave and is used for acquiring a scale mark image of the steel wire rope; the wireless transmission device is respectively connected with the image acquisition device and the control cabinet, and the wireless transmission device is used for wirelessly transmitting the scale mark image acquired by the image acquisition device and the revolution of the traction sheave motor to the monitoring center. By adopting the device and the method, the problem that elevator maintenance personnel are dangerous in the measurement process is solved.
Disclosure of Invention
The purpose of the invention is: and processing the input image data by using an image algorithm and a high-frame camera so as to obtain the movement amount of the steel wire rope of the elevator traction braking device.
In order to achieve the aim, the technical scheme of the invention provides a method for monitoring the movement amount of a steel wire rope in the braking process of an elevator, which is characterized by comprising the following steps of:
shooting a traction steel wire rope group in the braking process by using an industrial camera, transmitting the shot image into a computer, calculating the movement amount of the traction steel wire rope by using an image recognition algorithm by the computer, and displaying the result on a display device, wherein the image recognition algorithm comprises the following steps:
step 101, obtaining an image only containing a steel wire rope, and counting a gray level frequency distribution histogram of the image to obtain a vector x with a length of 2560；
102, converting an image obtained by an industrial camera in real time into a gray-scale image and carrying out Gaussian blur to obtain a de-noised gray-scale image I (x, y), and obtaining a frequency distribution histogram vector and a vector x of each point in the gray-scale image I (x, y)0The Euclidean distance between the two points is obtained by assigning the gray value of each point in the gray map I (x, y) to be 1 or 0, so as to obtain a binary map B for extracting the coordinate position of the steel wire rope0；
Step 103, calculating the mean value and standard deviation of all point coordinates on the x axis and the y axis, wherein the mean value and the standard deviation are respectively Mx、My、 σx、σyThen, there are:
x is to bemin、xmax、ymin、ymaxRespectively serving as the minimum value and the maximum value of an x axis and the minimum value and the maximum value of a y axis, so as to divide an area from the gray-scale image I (x, y), wherein the area is a detection window;
step 2, dividing the steel wire rope in the detection window obtained in the step 1 according to the root, removing the area which does not contain the steel wire rope, finally obtaining a group of small detection blocks, and calculating the index value related to the movement of the rope through the gray scale in the detection blocks, wherein the method comprises the following steps:
step 201, using hough transform to segment and rotate the detection window Sec (x, y) to obtain straight line parameters of all straight lines, wherein:
the straight line includes all wire rope and the straight line boundary between the wire rope in the detection window, interference straight line and redundant straight line, wherein: the straight line boundary is defined as a dividing line; the interference straight line is a straight line which is not parallel to the dividing line; the redundant straight line is parallel to the dividing line, but is close to another straight line, and can be regarded as a straight line of the same straight line;
the straight line parameters are the distance rho from the straight line to the origin and the included angle theta between the straight line and the x axis, and a set S is obtained through Hough transformation, wherein S is { (rho)i,θi)|i＝1,2,…,n}，ρiThe distance from the ith straight line to the origin; thetaiIs the included angle between the ith straight line and the x axis; n is the total number of detected straight lines
Step 202, processing the straight line parameters of all the straight lines detected in step 201, and eliminating interference straight lines and redundant straight lines, comprising the following steps:
step 2021, divide the set S, order
Any two of the sets Ai、AjMutual exclusion, the division is carried out according to the following method:
the parameter pair (rho) of the next straight line is taken out from the set Si,θi) And the included angle theta therebetweeniWith all existing sets A1、A2、…AkAverage number of all straight line angles in theta1mean、θ2mean、…θkmeanMaking a comparison if any one θ existssmeanSo that thetai-θsmean|＜TθIs established, TθIs a preselected threshold, the angle θ is further determinediCorresponding distance ρiAnd average number thetasmeanCorresponding set AsAverage value ρ of distances of all the straight lines insmeanBy contrast, if | ρi-ρsmean|＜TρIf true, then (ρ)i,θi) Join Collection As，TρIs a preselected threshold value, if | ρi-ρsmean|＜TρIf not, discarding the straight line;
if such an average number θ is not presentsmeanOr the existing set does not exist, a set A is newly establishedk+1And will (p)i,θi) Add New set Ak+1(ii) a Then continue to take down from the set SContinuing the above process until all the linear parameter pairs in the set S are completely obtained;
step 2022, through step 2021, the detection result of the hough transform is divided into a plurality of pairwise mutually exclusive sets a1、A2、…AmThe union of (1) is set as set A by finding out the one with the largest number of elements from the setsqThen set AqThe straight lines in the set are the straight lines which are correctly detected, namely the dividing lines, and the straight lines in other sets are all removed;
step 203, dividing the steel wire rope into independent areas according to the dividing lines obtained in the step 202, wherein each area is a detection zone, counting the pixel gray level in each detection zone, and calculating the obtained gray level distribution vector and the vector x0The Euclidean distance is calculated, the detection zones larger than the threshold value are removed, only the effective detection zones are reserved, and k effective detection zones are obtained;
step 204, setting I (x)1:x2,y1:y2) Representing the x-axis coordinate in the gray-scale image I (x, y) as x1To x2Y axis coordinate of y1To y2Then a valid detection band obtained in step 203 is represented as I (0: x)max,yi:yi+1) Wherein x ismaxMaximum x-axis coordinate representing the gray scale map I (x, y);
step 205, setting a block height h, the detection blocks of the detection zone in step 204 are I (0: h-1, y)i:yi+1)、I(1:h,yi:yi+1)、…、I(xmax-h+1:xmax,yi:yi+1) In total of xmax-h +2 detection blocks;
step 206, processing each effective detection zone by the method of step 204 and step 205 to obtain (x)max-h +2) k detection blocks;
step 207, summing each detection block obtained in step 206 along the x-axis direction to obtain sum (y), and counting the sum (y) in the middle coordinate ymidAnd ymid-1、ymidThe average value of +1 three points is used as an index value to countDrawing a line graph by using index values obtained by a detection block at different time;
and 3, calculating the effective fluctuation period number of the line graph changing along with time in each detection block, averaging, and multiplying the number by the strand spacing of the steel wire rope to obtain the moving distance of the steel wire rope.
Preferably, the industrial camera transmits the captured image to the computer using an industrial router.
Preferably, in step 102, for any point (I, j) in the gray scale map I (x, y), the points of a certain pixel region around the point are counted to obtain a frequency distribution histogram vector x of the gray scaleij(ii) a Calculating a frequency distribution histogram vector xijAnd vector x0And setting a threshold value TdIf dist (i, j) ═ x |, (i, j) | x) is satisfiedij-x0||2＜TdIf so, the gray value of the point (i, j) is assigned to 1, otherwise, the gray value of the point (i, j) is assigned to 0.
Preferably, the step 201 comprises the steps of:
step 2011, gradient threshold filtering is performed on the detection window Sec (x, y), which includes:
order:
obtaining a binary image Bedges(x, y) in the formulae (3) and (4), gradxSec is the gradient of the detection window along the x direction; gradySec is the gradient of the detection window along the y direction; mag (x, y) is a value after the gradient size of the detection window is regulated; t isMLIs the lowest threshold; t isMHIs the highest threshold;
step 2022, the binary image Bedges(x, y) performing Hough transform and detecting all the straight lines.
Preferably, the stepsIn step 203, set A is computedqAverage value theta of angles of all straight lines inqmeanThen the detection window is rotated counterclockwise by an angle thetaqmean；
Let set AqThe distances of the middle straight lines are respectively rho from small to largeq1＜ρq2＜…＜ρqmaxTaking [ rho ] for the y-axis of the imageq1+σ,ρq2-σ]、[ρq2+σ,ρq3-σ]、…、[ρq(max-1)+σ,ρqmax-σ]In these several intervals, each interval is a detection zone, where σ is a smaller preset value.
Preferably, in step 3, a threshold T is set for the amplitude of the fluctuation of the line graph during the detection period, and each time the function in the line graph increases beyond the threshold T and no decrease beyond the threshold T occurs midway, a new period is considered to occur, and the number of strands moved by the wire rope is increased by one.
The invention uses a high-frame industrial camera and an image recognition method to measure the movement amount of the steel wire rope. The image processing algorithm adopted by the invention can automatically identify the distribution positions of the steel wire ropes in the image and automatically segment the steel wire ropes. Therefore, the invention has low requirement on the installation of the camera and does not need to modify the traction braking device. The measurement based on the image is non-contact measurement, so that the defect of a friction small wheel measurement method can be overcome, and the method is suitable for long-term monitoring.
The invention realizes non-contact measurement, the measuring device does not influence the performance of the brake, the traction brake device does not need to be modified, the measured object can be found out according to the image, the requirement on the installation precision is reduced, the influence of the environment change on the detection system is also reduced, and the task of monitoring the movement of the steel wire rope for a long time can be completed.
Drawings
Fig. 1 is a schematic view of a traction brake apparatus in actual use;
FIG. 2 is a system diagram in the present embodiment;
FIG. 3 is a schematic diagram of a rectangular area obtained by manual cutting in the embodiment;
FIG. 4Is a binary image B of the coordinate position of the steel wire rope0；
FIG. 5 is a detection window;
FIG. 6 is a binary image Bedges；
FIG. 7 is a schematic view of a straight line drawn in the detection window;
FIG. 8 is a graph of the obtained waveforms;
fig. 9 is a process of calculating the elapsed period number using a waveform diagram.
Detailed Description
The invention will be further illustrated with reference to the following specific examples. It should be understood that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention. Furthermore, it should be understood that various changes and modifications can be made by those skilled in the art after reading the teachings of the present invention, and such equivalents also fall within the scope of the appended claims.
The invention provides a monitoring method for elevator steel wire rope movement amount based on image recognition, which inputs high frame rate video collected by an industrial camera (hereinafter referred to as a camera), and the steel wire rope texture change period is short, so that the steel wire rope texture change period can be completely collected by using the camera with high frame rate and short exposure time. In addition, in order to reduce the requirement on the placement position of the camera, the calculation parameters need to be updated according to the acquired data, and when the relative position of the camera and the brake wheel slightly changes, the camera can still calculate the movement amount of the steel wire rope.
The overall scheme of the invention is described in detail as follows:
the camera is placed at a proper position, and the camera transmits the collected image data to the industrial router and sends the image data to the computer through the industrial router. The computer runs an image processing program. After the image containing the steel wire rope is obtained, the points where the steel wire rope is located in the image need to be screened out, and the detection window containing the steel wire rope is divided according to the points. A plurality of steel wire ropes are arranged in the detection window, main straight lines in the detection window are detected by Hough transform, and the steel wire ropes are divided into a plurality of steel wire ropes by the straight lines to obtain a detection belt. The detection belt of each steel wire rope is further divided to obtain a plurality of small detection blocks, the period of each small detection block passing along the time change is counted, the average value is taken to obtain the number of strands of the steel wire rope moving, the slippage of the steel wire rope can be obtained by multiplying the number of strands by the distance of the strands, and finally the slippage is displayed on an output device (such as a display).
According to the foregoing description, the basic steps of the above-described image processing program are:
firstly, detecting a steel wire rope;
secondly, dividing the region;
and thirdly, calculating the movement amount.
The above-described image processing program is specifically described as follows:
(1) and (4) detecting the steel wire rope, namely detecting the coordinate distribution of points belonging to the steel wire rope in the image, thereby dividing an area only containing the steel wire rope to obtain, and preparing for the next calculation. The technique used is local histogram feature matching.
Compared with other parts of the image, the steel wire rope has higher gray scale change frequency along the spatial axis and certain periodicity, and belongs to a texture image. Such images cannot be simply segmented using color or gradient thresholding. Considering the regularity of spatial distribution of the texture image, a frequency histogram of gray levels in the surrounding area of each pixel point can be counted to form a vector with the length of 256. And comparing the vector with the gray level distribution vector of the area sampled in advance, and screening out the area where the steel wire rope is located.
Firstly, shooting any frame of image by using a camera, manually intercepting one section of image only containing a steel wire rope, counting a gray frequency distribution histogram of the image to obtain a vector x with the length of 2560。
And converting the original image into a gray image and carrying out Gaussian blurring to obtain a de-noised gray image I (x, y). For any point (I, j) in the gray scale image I (x, y), the points in the surrounding 13 × 13 pixel region are counted to obtain the frequency distribution histogram vector x of the gray scaleij. Computing a frequency distribution histogram vector xijAnd vector x0And setting a threshold valueValue TdI.e. when the formula:
dist(i,j)＝||xij-x0||2＜Td (1)
assigning the gray value of the satisfied point (i, j) to be 1, and assigning the rest points to be 0 to obtain a binary image B for extracting the coordinate position of the steel wire rope0。
Dividing an area which only contains the steel wire ropes as far as possible from the original image for analysis, and solving the mean value and standard deviation of all point coordinates on the x axis and the y axis, wherein the mean value and the standard deviation are respectively Mx、My、σx、σy. Order:
x is to bemin、xmax、ymin、ymaxAn area is divided from the gray scale map I (x, y) as the minimum and maximum values of x and the minimum and maximum values of y, and the area is a detection window, and the detection window may include other parts besides the steel wire ropes, for example, when the gap between the steel wire ropes is large, the gap may also include the detection window.
(2) And the area division is to divide the steel wire rope in the detection window obtained in the previous step according to the root, eliminate the area without the steel wire rope, finally obtain a group of small detection blocks, and calculate the index value related to the rope movement through the gray scale in the detection blocks. The region segmentation is divided into two steps, wherein the first step is to segment the detection window by using Hoff transformation and rotate the detection window. And (5) dividing the rotated detection window by utilizing the straight line detected by Hough transform to obtain a detection band. The second step is to extract the detection block from the detection band by using a sliding window method, and prepare for calculating the index value.
An obvious straight line boundary is formed between the steel wire rope and the steel wire rope in the detection window, and a long straight line in a graph can be detected by Hough transform and used as a dividing line. When Hough transform is used, gradient threshold filtering is firstly carried out on the detection window. The detection window is denoted by Sec (x, y). Order:
order:
obtaining a binary image Bedges(x, y) in the formulae (3) and (4), gradxSec is the gradient of the detection window along the x direction; gradySec is the gradient of the detection window along the y direction; mag (x, y) is a value after the gradient size of the detection window is regulated; t isMLIs the lowest threshold; t isMHIs the highest threshold.
The binary image B can then be processededges(x, y) are subjected to Hough transform, and all the straight lines are detected.
The segmentation lines obtained by the hough transform tend to be accompanied by some disturbing extraneous straight lines. All detected straight line parameters need to be processed, all straight lines which do not meet requirements are removed, and subsequent processing is facilitated. From the foregoing discussion, the lines that need to be culled fall into two categories:
1) an interfering straight line not parallel to the correct dividing straight line.
2) Parallel to the correct straight line of division, but very close to the other, can be regarded as redundant straight lines of the same straight line.
The hough transform returns a set S of the distance ρ from the straight line to the origin and the angle θ between the straight line and the x-axis:
S＝{(ρi,θi)|i＝1,2,…,n} (5)
in the formula (5), rhoiThe distance from the ith straight line to the origin; thetaiIs the included angle between the ith straight line and the x axis; n is the total number of detected lines.
According to the characteristics of the first two types of straight lines, a division is made on the set S, i.e. the order is made
Any two of the sets Ai、AjMutually exclusive. The division is carried out according to the following method:
the parameter pair (rho) of the next straight line is taken out from the set Si,θi) And the included angle theta therebetweeniWith all existing sets A1、A2、…AkAverage number of all straight line angles in theta1mean、θ2mean、…θkmeanMaking a comparison if any one θ existssmeanSo that:
|θi-θsmean|＜Tθ (6)
in the formula (6), TθIf it is a pre-selected threshold, the included angle theta is adjustediCorresponding distance ρiAnd the average number thetasmeanCorresponding set AsAverage value ρ of distances of all the straight lines insmeanAnd if:
|ρi-ρsmean|＜Tρ (7)
then will (ρ)i,θi) Join Collection AsIn the formula (7), TρIs a pre-selected threshold; otherwise the line is discarded. If such an average number θ is not presentsmeanOr if the existing set does not exist, a set A is newly establishedk+1And will (p)i,θi) Add New set Ak+1. And then, continuously taking the parameter pair of the next straight line from the set S, and continuing the process until all the parameter pairs of the straight lines in the set S are completely taken.
Through the above processes, the detection result of Hough transform is divided into a plurality of pairwise mutually exclusive sets A1、A2、…AmThe union of (a). The most numerous elements are found from these sets, assuming set AqThen set AqThe straight lines in the set are the straight lines which are correctly detected, and all the straight lines in other sets are removed. It is assumed here that the number of correctly detected lines is the largest, whereas erroneously detected lines are only a small number of lines and the difference in distance and angle is large.
The screened straight lines can already divide the steel wire rope into independent areas according to strips. But do notThese straight lines have a certain angle, which is not favorable for the subsequent calculation processing. Thus, set A is calculatedqAverage value theta of angles of all straight lines inqmeanThen the detection window is rotated counterclockwise by an angle thetaqmean. Thus, the originally detected dividing line with a certain slope becomes a vertical line. Assumption set AqThe distances of the middle straight lines are respectively rho from small to largeq1＜ρq2＜…＜ρqmax. Only the y-axis (horizontal axis) of the image needs to be taken as [ rho ]q1+σ,ρq2-σ]、 [ρq2+σ,ρq3-σ]、…、[ρq(max-1)+σ,ρqmax-σ]The intervals are detection bands. Wherein, σ is a smaller preset value, and the influence caused by the change of the edge can be reduced by using σ. And then, counting the pixel gray level in each detection zone, calculating the Euclidean distance between the obtained gray level distribution vector and the sample vector, eliminating the detection zone larger than the threshold value, and only keeping the effective detection zone.
After a plurality of valid detection zones are obtained, the movement amount can be calculated. The amount of movement is calculated, requiring the use of a detection zone to obtain a number of detection blocks. It is known from the foregoing discussion that denoised gray maps are denoted by I (x, y), now convention symbol I (x)1:x2,y1:y2) Representing the x-axis coordinate in I (x, y) as x1To x2Y axis coordinate of y1To y2The area of (a). Then a certain detection band may be denoted as I (0: x)max,yi:yi+1) Wherein x ismaxRepresents the maximum x-axis coordinate of I (x, y).
To obtain a detection block, a block height h needs to be set first. The detection blocks of the detection zone are I (0: h-1, y)i:yi+1)、I(1:h,yi:yi+1)、…、I(xmax-h+1:xmax,yi:yi+1) In total of xmax-h +2 detection blocks. If there are a total of k detection bands, then a total of (x) can be obtainedmax-h +2) k detection blocks. For each detection block, sum (y) is obtained by summing it along the x-axis direction. Statistics sum (y) at the middle coordinate ymidAnd ymid-1、ymidAt three points of +1As an index value. And counting index values obtained by the same detection block at different time, and drawing a line graph.
(3) The shift amount calculation is to calculate the shift amount by first calculating the number of effective fluctuation cycles of the time-varying waveform pattern in each detection block and averaging. And then multiplying the number by the strand spacing of the steel wire rope to obtain the moving distance of the steel wire rope.
The detection period has two difficulties: firstly, the peak and trough change of each period is large, and the peak and trough cannot be judged simply through a threshold value. Secondly, the waveform will fluctuate slightly due to interference, and the calculation period cannot be simply increased or decreased. Therefore, a threshold value can be set for the amplitude of the fluctuation, and each time the function increases and exceeds a threshold value T, and the decrease exceeding the threshold value T does not occur in the midway, a new period is considered to occur, and the number of the strands moved by the steel wire rope is increased by one.
The following description will be made by taking the monitoring of the movement of the wire rope during the braking process of the traction brake apparatus as shown in fig. 1 as an example.
The key point of the invention is that a scheme for detecting the movement amount of the steel wire rope in the elevator braking process is designed for the elevator traction braking device. The overall detection scheme is shown in figure 2. The method comprises the steps of using an industrial camera installed in advance to collect image data, receiving image input of the industrial camera by a computer, and transmitting the image collected by the industrial camera to the computer once. The computer completes all image processing and calculation tasks, and the display screen displays calculation results.
Aiming at the overall scheme, the invention provides the following image processing specific steps:
firstly, detecting a steel wire rope;
secondly, dividing the region;
and thirdly, calculating the movement amount.
And writing a program according to the steps, and running the program on a computer to detect the movement amount of the steel wire rope.
(1) Wire rope detection
After the industrial camera is installed, the camera shooting is controlled firstlyArranging a frame of photo, manually intercepting a rectangular area only containing the steel wire ropes, counting a gray histogram of the area to obtain a vector x as shown in figure 30. Vector x0Is a vector of length 256, in which case the histogram has an intercept set to 1 gray value (in practice, for simplicity of calculation, increasing the running speed, one can set the intercept to a greater number, for example 2 or 5, in this case 1 is temporarily used as the intercept). Then, the original image is converted into a gray-scale image, a local histogram is counted in a 13 × 13 pixel area of the image with each point as the center, and for an area close to the edge and incapable of taking the 13 × 13 pixel points, the minimum (maximum) coordinate is the boundary coordinate of the image (in the actual operation process, the area around each point of the image can not be counted, and the area can be counted at intervals of 3 points or 5 points, so that the main distribution position of the steel wire rope in the image can be obtained, the operation speed can be greatly improved, and each pixel point is counted in the embodiment). According to the steps described above, for each point a corresponding histogram vector and sample vector x are calculated0Calculating Euclidean distance, setting the threshold value to be 0.12, and obtaining a binary image B of the coordinate position of the steel wire rope0As shown in fig. 4. The range of the detection window is calculated based on this, and the detection window is divided from the gray scale as shown in fig. 5 (the coefficient multiplied by the standard deviation is set to 0.7).
(2) Region segmentation
Firstly, the image is converted into a gradient binary image, a threshold value is set 70, and a binary image B is obtainededgesAs shown in fig. 6. For the binary image BedgesStraight lines of length greater than 150 are detected using the hough transform and drawn in the detection window, as shown in fig. 7. And (3) dividing the straight line set by the method described above, taking the subset with the most elements to obtain effective straight lines, rotating the detection window according to the effective straight lines, and dividing the detection zone. And then selecting a valid detection zone, and obtaining a detection block by using a sliding window method.
(3) Calculation of movement amount
A line graph, i.e., a waveform graph, in which the index value changes with time is obtained for each detection block, as shown in fig. 8, which is a waveform graph of only the last stage time of braking, and only one stage time of one detection block is given because the whole process image is long. For this waveform, the elapsed cycles are calculated using the process shown in fig. 9, and the calculated shift amounts of all the detection blocks are averaged in strand pitch, i.e., the shift amount, to reduce the error and obtain the final shift amount.
Claims (6)
1. A method for monitoring the movement amount of a steel wire rope in the braking process of an elevator is characterized by comprising the following steps:
shooting a traction steel wire rope group in a braking process by using an industrial camera, transmitting the shot image into a computer, calculating the movement amount of the traction steel wire rope by using an image recognition algorithm by the computer, and displaying the result on a display device, wherein the image recognition algorithm comprises the following steps:
step 1, detecting the coordinate distribution of points belonging to a steel wire rope in an image obtained by an industrial camera in real time, thereby dividing an area only containing the steel wire rope, wherein the area is a detection window, and the method comprises the following steps:
step 101, obtaining an image only containing a steel wire rope, and counting a gray level frequency distribution histogram of the image to obtain a vector x with a length of 2560；
102, converting an image obtained by an industrial camera in real time into a gray-scale image and carrying out Gaussian blur to obtain a de-noised gray-scale image I (x, y), and obtaining a frequency distribution histogram vector and a vector x of each point in the gray-scale image I (x, y)0The Euclidean distance between the two points is obtained by assigning the gray value of each point in the gray map I (x, y) to be 1 or 0, so as to obtain a binary map B for extracting the coordinate position of the steel wire rope0；
Step 103, calculating the mean value and standard deviation of all point coordinates on the x axis and the y axis, wherein the mean value and the standard deviation are respectively Mx、My、σx、σyThen, there are:
x is to bemin、xmax、ymin、ymaxRespectively serving as the minimum value and the maximum value of an x axis and the minimum value and the maximum value of a y axis, so as to divide an area from the gray-scale image I (x, y), wherein the area is a detection window;
step 2, dividing the steel wire rope in the detection window obtained in the step 1 according to the root, removing the area which does not contain the steel wire rope, finally obtaining a group of small detection blocks, and calculating the index value related to the movement of the rope through the gray scale in the detection blocks, wherein the method comprises the following steps:
step 201, using hough transform to segment and rotate the detection window Sec (x, y) to obtain straight line parameters of all straight lines, wherein:
the straight line includes all wire rope and the straight line boundary between the wire rope in the detection window, interference straight line and redundant straight line, wherein: the straight line boundary is defined as a dividing line; the interference straight line is a straight line which is not parallel to the dividing line; the redundant straight line is parallel to the dividing line, but is close to another straight line, and can be regarded as a straight line of the same straight line;
the straight line parameters are the distance rho from the straight line to the origin and the included angle theta between the straight line and the x axis, and a set S is obtained through Hough transformation, wherein S is { (rho)i,θi)|i＝1,2,…,n}，ρiThe distance from the ith straight line to the origin; thetaiIs the included angle between the ith straight line and the x axis; n is the total number of detected straight lines
Step 202, processing the straight line parameters of all the straight lines detected in step 201, and eliminating interference straight lines and redundant straight lines, comprising the following steps:
step 2021, divide the set S, order
Any two of the sets Ai、AjMutual exclusion, the division is carried out according to the following method:
the parameter pair (rho) of the next straight line is taken out from the set Si,θi) And the included angle theta therebetweeniWith all existing sets A1、A2、…AkIn which all straight lines form an included angleMean number theta1mean、θ2mean、…θkmeanBy comparison, if any one θ exists thereinsmeanSo that | θi-θsmean|<TθIs established, TθIs a preselected threshold, the angle θ is further determinediCorresponding distance ρiAnd average number thetasmeanCorresponding set AsAverage value ρ of distances of all the straight lines insmeanBy contrast, if | ρi-ρsmean|<TρIf true, then (ρ)i,θi) Join Collection As，TρIs a preselected threshold value, if | ρi-ρsmean|<TρIf not, discarding the straight line;
if such an average number θ is not presentsmeanOr the existing set does not exist, a set A is newly establishedk+1And will (p)i,θi) Add New set Ak+1(ii) a Then, continuously taking the parameter pair of the next straight line from the set S, and continuing the process until all the parameter pairs of the straight lines in the set S are completely taken;
step 2022, through step 2021, the detection result of the hough transform is divided into a plurality of pairwise mutually exclusive sets a1、A2、…AmThe union of (1) is set as set A by finding out the one with the largest number of elements from the setsqThen set AqThe straight lines in the set are the straight lines which are correctly detected, namely the dividing lines, and the straight lines in other sets are all removed;
step 203, dividing the steel wire rope into independent areas according to the dividing lines obtained in the step 202, wherein each area is a detection zone, counting the pixel gray level in each detection zone, and calculating the obtained gray level distribution vector and the vector x0The Euclidean distance is calculated, the detection zones larger than the threshold value are removed, only the effective detection zones are reserved, and k effective detection zones are obtained;
step 204, setting I (x)1:x2,y1:y2) Representing the x-axis coordinate in the gray-scale image I (x, y) as x1To x2Y axis coordinate of y1To y2Then a valid detection band obtained in step 203 is represented as I (0: x)max,yi:yi+1) Wherein x ismaxMaximum x-axis coordinate representing the gray scale map I (x, y);
step 205, setting a block height h, the detection blocks of the detection zone in step 204 are I (0: h-1, y)i:yi+1)、I(1:h,yi:yi+1)、…、I(xmax-h+1:xmax,yi:yi+1) In total of xmax-h +2 detection blocks;
step 206, processing each effective detection zone by the method of step 204 and step 205 to obtain (x)max-h +2) k detection blocks;
step 207, summing each detection block obtained in step 206 along the x-axis direction to obtain sum (y), and counting the sum (y) in the middle coordinate ymidAnd ymid-1、ymidTaking the mean value of the +1 points as an index value, counting the index values obtained by the same detection block at different time, and drawing a line graph;
and 3, calculating the effective fluctuation period number of the line graph changing along with time in each detection block, averaging, and multiplying the number by the strand spacing of the steel wire rope to obtain the moving distance of the steel wire rope.
2. The method for monitoring the amount of movement of the wire rope during braking of the elevator according to claim 1, wherein the industrial camera transmits the captured image to the computer by using an industrial router.
3. The method as claimed in claim 1, wherein in step 102, for any point (I, j) in the gray scale map I (x, y), the points in a certain pixel region around the point are counted to obtain a histogram vector x of the frequency distribution of the gray scaleij(ii) a Computing a frequency distribution histogram vector xijAnd vector x0And setting a threshold value TdIf dist (i, j) ═ x |, (i, j) | x) is satisfiedij-x0||2＜TdIf so, the gray value of the point (i, j) is assigned to 1, otherwise, the gray value of the point (i, j) is assigned to 0.
4. The method for monitoring the movement amount of the steel wire rope in the braking process of the elevator as claimed in claim 1, wherein said step 201 comprises the steps of:
step 2011, gradient threshold filtering is performed on the detection window Sec (x, y), which includes:
order:
obtaining a binary image Bedges(x, y) in the formulae (3) and (4), gradxSec is the gradient of the detection window along the x direction; gradySec is the gradient of the detection window along the y direction; mag (x, y) is a value after the gradient size of the detection window is regulated; t isMLIs the lowest threshold; t isMHIs the highest threshold;
step 2022, the binary image Bedges(x, y) performing Hough transform and detecting all the straight lines.
5. The method for monitoring the amount of movement of steel wire ropes in the braking process of elevator as claimed in claim 1, wherein in step 203, set a is calculatedqAverage value theta of angles of all straight lines inqmeanThen the detection window is rotated counterclockwise by an angle thetaqmean；
Let set AqThe distances of the middle straight lines are respectively rho from small to largeq1＜ρq2＜…＜ρqmaxTaking [ rho ] for the y-axis of the imageq1+σ,ρq2-σ]、[ρq2+σ,ρq3-σ]、…、[ρq(max-1)+σ,ρqmax-σ]These several intervals, each interval being one of said detection zones, where σ isA smaller preset value.
6. The method for monitoring the amount of movement of the wire rope during braking of the elevator according to claim 1, wherein in the step 3, a threshold T is set for the amplitude of fluctuation of the line graph during the detection period, and when the function in the line graph increases beyond the threshold T and no decrease beyond the threshold T occurs halfway, a new period is considered to occur, and the number of strands moved by the wire rope is increased by one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011076684.5A CN112408153B (en) | 2020-10-10 | 2020-10-10 | Method for monitoring movement amount of steel wire rope in elevator braking process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011076684.5A CN112408153B (en) | 2020-10-10 | 2020-10-10 | Method for monitoring movement amount of steel wire rope in elevator braking process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112408153A true CN112408153A (en) | 2021-02-26 |
| CN112408153B CN112408153B (en) | 2022-06-24 |
Family
ID=74854874
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011076684.5A Active CN112408153B (en) | 2020-10-10 | 2020-10-10 | Method for monitoring movement amount of steel wire rope in elevator braking process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112408153B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114834989A (en) * | 2022-05-27 | 2022-08-02 | 中国特种设备检测研究院 | Elevator intelligent safety detection system |
| CN115339977A (en) * | 2022-09-16 | 2022-11-15 | 江西维毅机电设备有限公司 | Elevator traction wheel slippage detection device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011132010A (en) * | 2009-12-25 | 2011-07-07 | Meidensha Corp | Wire rope inspecting device |
| CN104239411A (en) * | 2014-08-12 | 2014-12-24 | 中国科学技术大学 | Color and position clustering and angular point detection-based detection method for grid-shaped radar |
| DE112015005003T5 (en) * | 2014-11-04 | 2017-08-03 | Mitsubishi Electric Corporation | ROPE DIAMETER MEASURING SYSTEM; SURFACE MEASUREMENT DEVICE, METHOD FOR MEASURING A SIZE DIAMETER AND PROGRAM |
| CN109242880A (en) * | 2018-08-24 | 2019-01-18 | 国网福建省电力有限公司漳州供电公司 | A kind of power line extraction method based on image procossing |
| CN110950213A (en) * | 2019-12-30 | 2020-04-03 | 暨南大学 | Method for measuring relative slippage of elevator steel wire rope |
-
2020
- 2020-10-10 CN CN202011076684.5A patent/CN112408153B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011132010A (en) * | 2009-12-25 | 2011-07-07 | Meidensha Corp | Wire rope inspecting device |
| CN104239411A (en) * | 2014-08-12 | 2014-12-24 | 中国科学技术大学 | Color and position clustering and angular point detection-based detection method for grid-shaped radar |
| DE112015005003T5 (en) * | 2014-11-04 | 2017-08-03 | Mitsubishi Electric Corporation | ROPE DIAMETER MEASURING SYSTEM; SURFACE MEASUREMENT DEVICE, METHOD FOR MEASURING A SIZE DIAMETER AND PROGRAM |
| CN109242880A (en) * | 2018-08-24 | 2019-01-18 | 国网福建省电力有限公司漳州供电公司 | A kind of power line extraction method based on image procossing |
| CN110950213A (en) * | 2019-12-30 | 2020-04-03 | 暨南大学 | Method for measuring relative slippage of elevator steel wire rope |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114834989A (en) * | 2022-05-27 | 2022-08-02 | 中国特种设备检测研究院 | Elevator intelligent safety detection system |
| CN115339977A (en) * | 2022-09-16 | 2022-11-15 | 江西维毅机电设备有限公司 | Elevator traction wheel slippage detection device |
| CN115339977B (en) * | 2022-09-16 | 2023-06-20 | 江西维毅机电设备有限公司 | Elevator traction sheave detection device that slides |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112408153B (en) | 2022-06-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325935B (en) | Power transmission line detection method based on unmanned aerial vehicle image | |
| CN102176228B (en) | Machine vision method for identifying dial plate information of multi-pointer instrument | |
| CN112408153B (en) | Method for monitoring movement amount of steel wire rope in elevator braking process | |
| CN102737370B (en) | Method and device for detecting image foreground | |
| CN110458157B (en) | Intelligent monitoring system for power cable production process | |
| CN107729853B (en) | Automatic identification method suitable for narrow-scale pointer instrument of transformer substation | |
| CN102236784A (en) | Screen area detection method and system | |
| CN114494210A (en) | Plastic film production defect detection method and system based on image processing | |
| CN102441581A (en) | Machine vision-based device and method for online detection of structural steel section size | |
| CN107742307A (en) | Based on the transmission line galloping feature extraction and parameters analysis method for improving frame difference method | |
| CN115861291A (en) | Patch circuit board production defect detection method based on machine vision | |
| CN115311629B (en) | Abnormal bending precision monitoring system of bending machine | |
| CN116402863B (en) | Intelligent analysis and early warning system for building construction monitoring data | |
| CN113155839A (en) | Steel plate outer surface defect online detection method based on machine vision | |
| CN116630813B (en) | Highway road surface construction quality intelligent detection system | |
| CN115311267B (en) | Method for detecting abnormity of check fabric | |
| CN116071657B (en) | Intelligent early warning system for building construction video monitoring big data | |
| CN116664569B (en) | Weld flash defect detection method | |
| CN116385450A (en) | PS sheet wear resistance detection method based on image processing | |
| CN107516315B (en) | Tunneling machine slag tapping monitoring method based on machine vision | |
| CN115841633A (en) | Power tower and power line associated correction power tower and power line detection method | |
| CN111127381B (en) | Non-parallel detection method for pantograph slide plate | |
| CN116110006B (en) | Scenic spot tourist abnormal behavior identification method for intelligent tourism system | |
| CN113252103A (en) | Method for calculating volume and mass of material pile based on MATLAB image recognition technology | |
| CN117274293A (en) | Accurate bacterial colony dividing method based on image features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |