CN110987383B

CN110987383B - An image-based measurement method of high-frequency two-position reciprocating action state

Info

Publication number: CN110987383B
Application number: CN201911313677.XA
Authority: CN
Inventors: 袁嫣红; 曾洪铭
Original assignee: Zhejiang University of Technology ZJUT
Current assignee: Zhejiang University of Technology ZJUT
Priority date: 2019-12-19
Filing date: 2019-12-19
Publication date: 2021-07-13
Anticipated expiration: 2039-12-19
Also published as: CN110987383A

Abstract

The invention relates to a method for measuring a high-frequency two-position reciprocating motion state based on an image, and belongs to the technical field of image processing. The motion trail of a high-speed reciprocating motion object is obtained by prolonging the exposure time, and the motion trail is subjected to gray level processing and further subjected to differential processing with a standard image. And analyzing the reliability of the score of the movement of the object which moves back and forth at high speed between the two polar positions according to the obtained difference result.

Description

Image-based measuring method for high-frequency two-position reciprocating motion state

Technical Field

The invention relates to a method for measuring a high-frequency two-position reciprocating motion state based on an image, and belongs to the technical field of image processing.

Background

The object motion detection technology is one of the basic technologies of modern industry and is the key for ensuring the product quality. With the improvement and improvement of industrial manufacturing technology and processing technology, higher requirements are put on detection means, detection speed and precision, and therefore, the traditional contact type measurement method cannot meet the requirements. Therefore, non-contact measurement methods based on photoelectric, electromagnetic, and other technologies are emerging. With the rapid development of machine vision technology, non-contact measurement means based on image detection are widely used in various fields. The vision measurement technology uses the image as a means or carrier for detecting and transmitting information, extracts useful signals from the image, and obtains various required parameters by processing the detected image, and has the characteristics of non-contact, full-field measurement, high precision and high automation degree. However, at present, the visual measurement of the reciprocating motion of a high-speed moving object is usually tested by an expensive high-frequency industrial camera, and the required cost is high. In order to avoid the use of a high frame rate camera and reduce the measurement cost, the invention hopes to provide a method for measuring the high-frequency two-position reciprocating motion state based on an image.

Disclosure of Invention

In order to avoid using a high frame rate camera when measuring the reciprocating motion of a moving object, thereby reducing the cost, the invention provides a method for measuring the high-frequency two-position reciprocating motion state based on an image, which has the following specific scheme:

a method for measuring a high-frequency two-position reciprocating motion state based on an image comprises the following steps:

(1) taking two industrial cameras, namely a first industrial camera and a second industrial camera, adjusting the placing positions and angles of the two cameras to enable the moving part to move within the field range of the two industrial cameras and enable the cameras to be focused on the whole moving track of the object to be measured;

(2) setting the exposure time of the two industrial cameras as integral multiple of the time required by the moving object to carry out one reciprocating motion, wherein the two industrial cameras adopt a pulse width triggering mode and a non-overlapping working mode in the testing process;

(3) the first industrial camera starts exposure operation when receiving a rising edge trigger signal; stopping exposure operation when the first industrial camera receives a falling edge trigger signal, and starting exposure operation by the second industrial camera; stopping exposure operation when the second industrial camera receives the rising edge trigger signal, starting exposure operation of the first industrial camera at the same time, and starting the next exposure cycle; the obtained image data is transmitted back to the computer for processing;

(4) sampling to obtain the round-trip motion of the moving object for integer times, and carrying out gray level processing and light intensity normalization processing on the image. Firstly, collecting an image of a moving part, converting the image into a Gray image according to a weighting method, calculating the Gray image such as Gray ═ R + G + B)/3 to obtain a Gray value F (x, y) of the image, and taking the Gray value F (x) of a fixed mark point on the image in order to eliminate the influence of the time variation of the light source intensity on the Gray value of the image₀,y₀)，F_g(x,y)＝F(x,y) /F(x₀,y₀) Namely the normalized image gray scale. The gray value of the normalized standard image for comparison obtained in normal operation is recorded as B_g(x, y), the gray value of the nth frame image obtained in the test is f_ng(x, y). In order to reduce the amount of computation, the measured image is compared with the standard image, and only the images of the two extreme position regions need to be compared. The comparisons described below are therefore all made for two extreme position areas. The specific comparison method is as follows:

will f is_ng(x, y) and the standard image gray value B_g(x, y) are subtracted to obtain a difference result D_n(x, y); one by oneEach pixel point is subjected to binarization processing to obtain a binarization image R_n(ii) a When D is present_n(x,y)>At T-delta, R_n(x, y) 255, when D_n(x,y)<At T-delta, R_n(x, y) is 0, wherein the threshold value T is 1/N of the gray value of a certain pixel point at the extreme position of the standard image, namely T is B_g(x, y)/N, delta is an allowable deviation; and (3) taking a pixel with the numerical value of 255 in the binary image as a foreground target pixel and a pixel with the numerical value of 0 as a background pixel, and judging the swing condition of the moving object according to the distribution condition of the extreme position pixels in the binary image.

And the time when the moving part starts to move in the third step is just the time when the first industrial camera starts to expose, and the setting can ensure that an integral number of times of moving part imaging pictures can be sampled after one exposure.

And in the third step, the first industrial camera and the second industrial camera transmit data back to the computer for processing when exposure operation is not performed, the setting can ensure that meter leakage is avoided, and the cameras can acquire information at any time.

The first industrial camera acquires odd frame images, and the second industrial camera acquires even frame images.

The exposure time of the two industrial cameras is less than or equal to the time interval between each frame. The setting can avoid image coincidence between frames, and the industrial camera can output complete pictures.

The difference in the fourth step is inversely related to reliability. The larger the difference, the lower the reliability of the reciprocating motion of the moving object.

The principle of the invention is as follows:

a camera with a common frame rate is adopted, an integral number of times of swing component imaging pictures are sampled after one exposure, the motion trail of an object which moves at a high speed between two polar positions is recorded, the gray scale processing is carried out on the object, the obtained image gray scale value is subjected to the difference processing with the standard gray scale value under the normal operation (namely, the object can reach the two polar positions each time), and the larger the difference value at the end point position is, the lower the reliability of the reciprocating motion of the moving object is.

Has the advantages that:

1. for a moving object which performs high-frequency reciprocating motion in the measurement, a high-frame-rate camera is not needed, so that the measurement cost is reduced;

2. compared with the traditional detection means and method, the method provided by the invention has the advantages of high efficiency, high reliability and high universality.

Drawings

FIG. 1 is a schematic view of alternate operation of two industrial cameras;

fig. 2 is a flowchart of performing image post-processing.

Detailed Description

The following is further explained with reference to the drawings of the specification.

A method for measuring the high-frequency two-position reciprocating motion state based on an image is used for testing the reliability of the reciprocating swing of a cutter head of a needle selector in a circular knitting machine at two limit positions.

A needle selector swing measuring platform is built and mainly comprises a tested electromagnetic needle selector, a three-degree-of-freedom micro-displacement adjusting platform, a control circuit, a voltage-stabilized power supply and two industrial cameras with the model number of G3-GM 10-C1240. The method comprises the steps that two industrial cameras are named as a first industrial camera and a second industrial camera respectively, exposure trigger modes of the first industrial camera and the second industrial camera are set to be pulse width trigger modes, and working modes of the two industrial cameras are selected to be non-overlapping working modes. To avoid coincidence between frames due to too long exposure time, the exposure time of the first industrial camera was set to 1/50s and the exposure time of the second industrial camera was set to 1/50 s. The tested needle selector is fixed on the platform, and the three-degree-of-freedom micro-displacement adjusting platform is adjusted, so that the motion trail of the tool bit of the tested needle selector is within the effective detection range of the first industrial camera and the second industrial camera. The output of the stabilized voltage power supply is connected with a control circuit, and the control circuit outputs excitation voltage (the normal working voltage of the needle selector is generally 24V) to drive the needle selector to swing back and forth at the two limit positions. The needle selector tool bit swings from the left side to the right side under the driving of the forward pulse voltage, and the starting time of the swing of the needle selector tool bit is the sending time of the driving pulse. When the needle selector tool bit receives the driving pulse signal, a rising edge trigger signal of the pulse width trigger signal is given to the first industrial camera, so that the first industrial camera starts exposure at the swing starting moment of the needle selector. The first industrial camera starts exposure operation when receiving a rising edge trigger signal of the pulse width trigger signal and stops exposure operation when receiving a falling edge trigger signal; when the first industrial camera stops exposure, the second industrial camera starts exposure operation; and stopping the exposure operation when the second industrial camera receives the rising edge trigger signal, starting the exposure operation of the first industrial camera and starting the next exposure cycle. The first industrial camera and the second industrial camera are exposed in turn, so that the image data stored in the memory can be transmitted to the computer for image data processing when the first industrial camera and the second industrial camera are not exposed, and simultaneously, the collected images can be ensured as many as possible, and the working efficiency is improved.

When the fatigue test is performed on the needle selector under the driving voltage of 24V, the swing frequency of the cutter head reaches 1000 times/second, and the exposure time of the two cameras is 1/50s, so that each frame image comprises 20 motion tracks, namely N is 20. For each image, converting the color image into a Gray image according to the formula Gray ═ R + G + B)/3, and performing light source standardization processing on each pixel, namely dividing each pixel by the Gray value of the fixed mark point in the image, which is called a normalized image. And averaging the gray scale of each pixel point in the 100 frames of normalized images initially acquired by the needle selector to obtain an initial standard image. And comparing the 100 frames of images with the initial standard image to remove the 20 frames of images with large deviation. The remaining 80 normalized images were re-averaged to be the standard image for the actual measurement. Each frame of image in the experimental test is also converted into a normalized gray image according to the formula. Taking the gray value of each pixel point in the area around the limit position (the width is one tenth of the track width) to make difference, and setting the x coordinate of the pixel point at the left limit position as 0 and the x at the right limit position as x_Pole(s)The coordinates are given as if for 0<|x-x _Pole(s)|<x_Pole(s)/20，0<|x|<x_Pole(s)Each 20 satisfies f_ng(x,y)-B_g(x,y)＝D_nIf (x, y) ═ 0, it means that the motion trajectory in the frame image is normal. If one of the conditions is not satisfied, the description is givenThe moving part swings abnormally. Recording the gray value difference D between the abnormal image and the standard image_n(x, y), and continuing to perform binarization processing on the abnormal image. Selecting proper deviation value delta according to measurement requirements>0, if D_n(x,y)>B_g(x,y)/20-δ，R_n(x, y) 255, when D_n(x,y)<B_g(x, y)/20-delta, R_n(x, y) ═ 0. R in the binary image obtained finally_nThe pixel distribution area with (x, y) ═ 255 is the swing abnormal area. If 0<|D_n(x,y)-B_g(x,y)/20|<δ, meaning not swung into place once, 0<|D_n(x,y)-2*B_g(x,y)/20|<δ, representing not swinging into position 2 times, 0<|D_n(x,y)-3*B_g(x,y)/20|<δ, which represents not swinging into position 3 times, and so on. And recording the times of not swinging to the right position, wherein the more times, the worse the reliability is.

The above-mentioned embodiments are only used for explaining the inventive concept of the present invention, and do not limit the protection of the claims of the present invention, and any insubstantial modifications of the present invention using this concept shall fall within the protection scope of the present invention.

Claims

1. a measurement method based on the high-frequency two-position reciprocating action state of an image, is characterized in that, comprises the following steps:

(1) Take two industrial cameras, namely the first industrial camera and the second industrial camera, and adjust the placement position, angle and focal length of the two cameras, so that the moving parts move within the field of view of the two industrial cameras;

(2) The exposure time of the two industrial cameras is set to an integer multiple of the time required for a moving object to perform a round-trip motion. During the test, the two industrial cameras both use the pulse width trigger mode and the non-overlapping working mode;

(3) The exposure operation starts when the first industrial camera receives the rising edge trigger signal; the exposure operation stops when the first industrial camera receives the falling edge trigger signal, and the second industrial camera starts the exposure operation at the same time; the second industrial camera receives the rising edge When the trigger signal is triggered, the exposure operation is stopped, and at the same time, the first industrial camera starts the exposure operation and starts the next exposure cycle;

(4) Sampling to obtain the round-trip motion exposure image of an integer number of moving objects, subtract the gray value f _ng (x, y) of the nth frame image from the gray value B _g (x, y) of the standard image, and obtain the difference result D _n (x, y); binarize the pixels one by one to obtain a binarized image R _n ; when D _n (x, y)>T-δ, R _n (x, y)=255, when When D _n (x, y)<T-δ, R _n (x, y)=0, where the threshold T is 1/N of the gray value of a pixel at the limit position of the standard image, that is, T=B _g (x , y)/N, δ is the allowable deviation; the pixel with a value of 255 in the binarized image is used as the foreground target pixel, and the pixel with a value of 0 is used as the background pixel. According to the pixel distribution at the extreme position in the binary image, the moving object is judged. Swing situation; the N is the number of motion tracks.

2 . The method for measuring a high-frequency two-position reciprocating action state based on an image according to claim 1 , wherein the moment when the third-step moving part starts to move is exactly the moment when the first industrial camera starts to expose. 3 .

3 . The method for measuring a high-frequency two-position reciprocating action state based on an image according to claim 1 , wherein in the third step, the first industrial camera and the second industrial camera are not performing exposure operations. 4 . , transfer the data back to the computer for processing.

4 . The image-based measuring method of high-frequency two-position reciprocating action state according to claim 1 , wherein the exposure time of the two industrial cameras is less than or equal to the time interval between each frame. 5 .

5 . The method for measuring high-frequency two-position reciprocating motion state based on an image according to claim 1 , wherein the first industrial camera collects odd-numbered frame images, and the second industrial camera collects even-numbered frame images. 6 .

6 . The method for measuring the high-frequency two-position reciprocating action state based on an image according to claim 1 , wherein the difference in the fourth step is negatively correlated with reliability. 7 .