Search

CN-121724990-B - Paint mist stable state change judging method based on image sequence cross correlation

CN121724990BCN 121724990 BCN121724990 BCN 121724990BCN-121724990-B

Abstract

A paint mist steady state change judging method based on image sequence cross-correlation relates to the technical field of image data processing and aims to solve the problems that in the prior art, a paint mist cone angle is calculated only through a single frame of static image, paint mist dynamic change cannot be reflected, the paint mist is easy to be interfered by accidental factors and a stable reference is lacking. According to the invention, by collecting the paint mist image sequence and generating the multi-frame average reference image, the traditional single-frame image detection mode is replaced, accidental interference of single-frame data is effectively avoided, a stable and reliable paint mist standard reference state is established, and the anti-interference capability and the result reliability of detection are greatly improved.

Inventors

  • SU CHENGZHI
  • YUE YUXIN
  • BAO HAIFENG
  • WANG ENGUO

Assignees

  • 长春理工大学

Dates

Publication Date
20260508
Application Date
20260225

Claims (8)

  1. 1. The paint mist stable state change judging method based on image sequence cross correlation is characterized by comprising the following steps of: step one, paint mist image acquisition; Step two, establishing a reference image; Carrying out arithmetic average calculation on the paint mist image sequence acquired in the first step to obtain an average image, and taking the average image as a reference image; Step three, image preprocessing; step four, calculating a fog cone angle; Step five, calculating the relative error of the fog cone angle; step six, calculating a cross-correlation coefficient; calculating to obtain normalized cross-correlation coefficients of the real-time image and the reference image; And seventhly, comprehensively judging the change of the stable state of the paint mist.
  2. 2. The method for judging the stable state change of paint mist based on the cross correlation of image sequences according to claim 1, wherein the first step is that a camera is installed at a proper position on one side of a spraying device, light is transmitted from one side of a mist cone to the other side, proper shooting frame rate and resolution are set, and when the spraying device is debugged to an optimal technological parameter, the method is characterized in that the method comprises the following steps of Frame gray paint mist image to obtain gray image sequence Wherein 。
  3. 3. The method for judging the stable state change of paint mist based on the image sequence cross correlation according to claim 2, wherein the step two is specifically to use a formula For grey image sequences Arithmetic average calculation is carried out to obtain an average image, and the average image is taken as a reference image ; 。
  4. 4. The paint mist steady state change judging method based on image sequence cross correlation according to claim 3, wherein the third step specifically comprises the following steps: Step 3.1, using OTSU algorithm to reference image Binarization processing is carried out to obtain a binary image The OTSU algorithm calculates based on the image histogram, calculates a threshold value when the variance among the image histogram classes is maximum, and segments out foreground and background images, so that the main content of the images is more prominent, and subsequent analysis and processing are facilitated; step 3.2 using the formula For binary images Performing morphological open operation processing to obtain a denoising image The morphological opening operation comprises a corrosion-before-expansion process; ; Formula (VI) In the process, the Is a structural element.
  5. 5. The paint mist steady state change judging method based on image sequence cross correlation according to claim 4, wherein the fourth step specifically comprises the following steps: step 4.1, denoising the image by using the Canny algorithm Edge extraction is carried out to obtain an edge image of the fog cone ; Step 4.2, utilizing Hough Transform algorithm to make edge image Performing linear detection to obtain a linear equation of the upper edge of the fog cone angle And fog cone angle lower edge linear equation , And Respectively represent the slope and intercept of the edge line on the fog cone angle, And Respectively representing the slope and intercept of the straight line at the lower edge of the fog cone angle; Step 4.3 utilizing the formula Calculating to obtain a reference image Is of the fog cone angle ; 。
  6. 6. The paint mist steady state change judging method based on image sequence cross correlation according to claim 5, wherein the fifth step specifically comprises the following steps: Step 5.1, collecting a frame of gray paint mist image at any time, and taking the frame of image as a real-time image ; Step 5.2 for real-time image Repeating the third step and the fourth step in sequence to obtain a real-time image Is of the fog cone angle ; Step 5.3, calculating by using the formula (4) And (3) with Differences of (2) ; (4)。
  7. 7. The method for determining the stable state change of paint mist based on the cross correlation of image sequences according to claim 6, wherein the step six is specifically as follows: calculating by using the formula (5) to obtain a real-time image And reference image Normalized cross-correlation coefficient of (2) ; And Respectively representing images And In coordinates of The gray value of the pixel at that point, , , Representing an image And Is defined by a set of pixel coordinates; (5); In the formula (5) of the present invention, And Respectively representing images And Average of all pixel gray values.
  8. 8. The paint mist steady state change judging method based on image sequence cross correlation according to claim 7, wherein the step seven is specifically: setting based on historical data Is of the threshold value of , Is of the threshold value of If (1) And is also provided with And judging that the stable state of the paint mist is unchanged, otherwise, judging that the stable state of the paint mist is changed.

Description

Paint mist stable state change judging method based on image sequence cross correlation Technical Field The invention relates to the technical field of image data processing, in particular to a paint mist steady state change judging method based on image sequence cross correlation. Background In industrial spraying operations, the stable state of the paint mist plays a decisive role in the quality of the coating. At present, a paint mist steady state change judging method mainly depends on manual experience and traditional physical measurement means. The manual evaluation mode has obvious defects, has strong subjectivity and large judgment standard difference of different operators, and causes lack of consistency and accuracy of evaluation results. And the manual evaluation efficiency is low, the real-time monitoring of the spraying process cannot be realized, and abnormal changes of the stable state of paint mist are difficult to discover in time. Although the traditional physical measurement methods can obtain partial parameters of paint mist to a certain extent, the methods can only perform single-point measurement generally and cannot comprehensively reflect the dynamic change condition of the paint mist in the whole spraying area. In addition, the traditional measuring method is complex to operate, professional equipment and personnel are needed, and production cost and measuring difficulty are increased. With the development of machine vision technology, it is possible to judge the stable state change of paint mist by collecting a paint mist image sequence and utilizing a digital image processing technology. A spray nozzle fog cone angle measuring method and device based on digital image processing is disclosed in Chinese patent publication No. CN109978905A, and the method comprises the steps of firstly preprocessing an acquired original image to obtain a binary image, then carrying out linear detection on the binary image to obtain a linear edge of a fog cone angle, and finally extracting a linear slope according to the linear edge to detect the size of the fog cone angle. Because the spraying operation is a dynamic process, the method mentioned in the patent only captures a static picture of a certain frame in the spraying process, and only processes the image of the frame to obtain the size of the fog cone angle for judging the stable state change of the paint fog. The method cannot reflect the dynamic change of paint mist, and when the mist cone angle is calculated, a single frame image is easily influenced by accidental factors, so that the accuracy and reliability of a calculation result can be influenced. More importantly, the existing method lacks a stable and reliable reference state as a reference, and cannot effectively judge the change of the current paint mist stable state, so that process feedback control is difficult to realize. Disclosure of Invention The invention provides a paint mist stable state change judging method based on image sequence cross correlation, which aims to solve the problems that in the prior art, a mist cone angle is calculated only through a single frame of static image, dynamic change of paint mist cannot be reflected, the paint mist is easy to interfere by accidental factors and a stable reference is lacking. The technical scheme of the invention is as follows: A paint mist stable state change judging method based on image sequence cross correlation comprises the following steps: step one, paint mist image acquisition; Step two, establishing a reference image; Step three, image preprocessing; step four, calculating a fog cone angle; Step five, calculating the relative error of the fog cone angle; step six, calculating a cross-correlation coefficient; And seventhly, comprehensively judging the change of the stable state of the paint mist. The first step is that a camera is arranged at a proper position on one side of a spraying device, lamplight is transmitted from one side of a fog cone to the other side, proper shooting frame rate and resolution are set, the spraying process is adjusted to meet the spraying requirement, and at the moment, the light is collectedFrame gray paint mist image to obtain gray image sequenceWherein。 The second step is to use the formula (1) to sequence gray scale imagesArithmetic average calculation is carried out to obtain an average image, and the average image is taken as a reference image; (1)。 The third step specifically comprises the following steps: Step 3.1, using OTSU algorithm to reference image Binarization processing is carried out to obtain a binary imageThe OTSU algorithm calculates based on the image histogram, calculates a threshold value when the variance among the image histogram classes is maximum, and segments out foreground and background images, so that the main content of the images is more prominent, and subsequent analysis and processing are facilitated; step 3.2, using equation (2) to apply to the binary imag