CN113808052B - Method for monitoring well cleaning in real time based on machine vision - Google Patents

Abstract

The invention relates to the technical field of drilling construction, and discloses a method for real-time monitoring of wellbore cleaning based on machine vision, which includes the following steps: S101: Collect images of rock cuttings on the vibrating screen surface; S102: Perform image contrast enhancement processing on the collected images; S103: Grayscale and binarize the contrast-enhanced image; S104: Perform cutting edge detection and morphological processing on the grayscale and binarized image; S105: Detect the cutting edge Perform smoothing filtering and pixel marking on the morphologically processed image; S106: Calculate the area of cuttings in the image, calculate the concentration of returned mud cuttings, and determine whether the wellbore is clean. The present invention calculates the area of cuttings and estimates the concentration of returned mud cuttings to determine whether the wellbore is clean. The present invention uses image recognition to monitor whether the wellbore is clean in real time. It is a non-contact measurement method that is safer and more reliable.

Description

A method for real-time monitoring of wellbore cleaning based on machine vision

Technical field

The invention relates to the technical field of drilling construction, and specifically discloses a method for real-time monitoring of wellbore cleaning based on machine vision.

Background technique

During the drilling construction process, due to the long borehole length and excessive final well inclination angle during operation of the extended reach well, cuttings are easily deposited in the highly inclined and horizontal well sections, forming a cuttings bed, causing the wellbore to be unclean. , causing repeated cutting of the drill bit, causing stuck pipes, wellbore blockage, formation rupture, well leakage, etc. Therefore, when the wellbore is not clean, we must promptly discover and take countermeasures.

In the existing technology, the method of weighing cuttings is generally used in engineering to calculate the density of cuttings in the annulus to determine whether the wellbore is clean. However, this method cannot monitor the cleanliness of the wellbore in real time, and the weighing process is cumbersome and inefficient. . In order to overcome the above problems, a method for real-time monitoring of wellbore cleanliness based on machine vision is provided. Machine vision is used to monitor the screen surface status of the vibrating screen to determine whether the wellbore is clean and to monitor the cleanliness of the wellbore in real time.

Contents of the invention

The present invention is intended to provide a method for real-time monitoring of wellbore cleaning based on machine vision, and to monitor whether the wellbore is clean in real time through machine vision and image recognition technology.

In order to achieve the above objectives, the basic solution of the present invention is as follows: a method for real-time monitoring of wellbore cleaning based on machine vision, including the following steps:

S101: Collect images of cuttings on the vibrating screen surface;

S102: Perform image contrast enhancement processing on the collected images;

S103: Grayscale and binarize the contrast-enhanced image;

S104: Perform cutting edge detection and morphological processing on the grayscale and binarized images;

S105: Perform smoothing filtering and pixel labeling on the image after edge detection and morphological processing of cuttings;

S106: Calculate the area of cuttings in the image, calculate the concentration of returned mud cuttings, and determine whether the wellbore is clean.

Further, in S101, the time interval for collecting cuttings images is T/2, where T is the conveyor belt period.

Further, in S102, in the process of enhancing the contrast of the cuttings image, the gamma value is set to 0.2 to enhance the brightness.

Further, in S103, the threshold value is selected as 0.4 during binarization processing to reduce image distortion.

Further, in S104, after cutting edge detection, the strel function is used to fill the gaps, in which a square structural element is used and the width is set to 3 pixels.

Further, in S106, for the processed cuttings image, the calculation formula for the area of each cuttings is as follows:

In the formula, S ₁ is the area of the cuttings on each image, P ₁ is the total pixels of the measured cuttings, P ₂ is the total pixels of the reference square, S ₀ is the area of a single reference square, and n is the number of reference squares.

Furthermore, after calculating the area of each cutting, estimate the mud concentration on the vibrating screen:

In the formula, S ₁ is the area of cuttings on each image, S is the area of each image, and a is the screening efficiency of the vibrating screen;

If η>5%, the wellbore is judged to be clean; if η<5%, the wellbore is judged to be unclean.

Principles and beneficial effects of the present invention: This solution makes full use of the high efficiency and high precision of machine vision to collect images of cuttings on the vibrating screen surface, calculate the content of the returned cuttings, and at the same time calculate the returned mud. Cuttings concentration is used to determine whether the wellbore is clean. It is a non-contact measurement method that is safer and more reliable than the mass flow meter method of measuring solid phase. Compared with the cuttings weighing method, it is faster and can monitor the cleanliness of the wellbore in real time.

Of course, implementing the applied solution does not necessarily require achieving all the above-mentioned technical effects at the same time.

Description of drawings

Figure 1 is a flow chart in an embodiment of the present invention;

Figure 2 is a schematic diagram of cuttings in an embodiment of the present invention;

Figure 3 is an enhanced contrast diagram of cuttings images in an embodiment of the present invention;

Figure 4 is a grayscale image of the cuttings image in the embodiment of the present invention;

Figure 5 is a binarized image of cuttings images in the embodiment of the present invention;

Figure 6 is a diagram of cutting edge extraction in the embodiment of the present invention;

Figure 7 is a diagram of cuttings gap filling in an embodiment of the present invention;

Figure 8 is a diagram of clear cuttings filling in the embodiment of the present invention;

Figure 9 is a cutting filter diagram in the embodiment of the present invention.

Detailed ways

The technical solutions in 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. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments.

In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inner", " The orientation or positional relationship indicated by "outside" and so on is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation. Specific orientations of construction and operation are therefore not to be construed as limitations of the invention.

Example:

Basically as shown in Figure 1: A method of real-time monitoring of wellbore cleaning based on machine vision, including the following steps:

S101: Collect images of cuttings on the vibrating screen surface;

S102: Perform image contrast enhancement processing on the collected images;

S103: Perform grayscale and binarization processing on the contrast-enhanced image;

S104: Perform cutting edge detection and morphological processing on the grayscale and binarized image;

S105: Perform smoothing filtering and pixel labeling on the image after edge detection and morphological processing of cuttings;

S106: Calculate the area of cuttings in the image, calculate the concentration of returned mud cuttings, and determine whether the wellbore is clean.

The specific implementation is as follows:

S101: Install the industrial explosion-proof camera directly above the screen surface of the vibrating screen, and collect images of cuttings on the screen surface of the vibrating screen with a time interval of T/2 and T as the conveyor belt transmission period, as shown in Figure 2, various shapes The cuttings are unevenly distributed on the vibrating screen.

S102: Preprocess the collected screen surface image of the vibrating screen. As shown in Figure 3, use the Imadjust function to enhance the contrast of the image. In the process of enhancing the contrast of the image, set the gamma value to 0.2 to enhance the brightness and make the image The contrast between the cuttings and other areas is clearer.

S103: Perform grayscale and binarization processing on the contrast-enhanced cuttings image. During the binarization process, the threshold is selected as 0.4, which can better reduce the distortion of the image. The results of image grayscale and binarization As shown in Figure 4 and Figure 5.

S104: Then perform edge detection on the processed cuttings image, select the canny operator for edge detection, locate all edge points, reduce the error rate, and make the positioning edge closer to the real edge of the cuttings. The edge extraction results of the cuttings image are as follows As shown in Figure 6.

After edge extraction of the cuttings image, the image morphology is processed, and the strel function is used to fill the gaps. Square structural elements are used and the width is set to 3 pixels to complete the gap filling. The filled cuttings image is shown in Figure 7 shown. Increase the width of the edge of the cuttings to close the unclosed areas to facilitate filling of the image.

Then fill the interior of the cuttings to obtain a complete cuttings image as shown in Figure 8, which facilitates subsequent pixel marking of the cuttings image and calculation of the area of the cuttings.

S105: Perform adaptive Gaussian filtering on the filled image to remove Gaussian noise in the image. First, a two-dimensional Gaussian filter function is used to generate a Gaussian kernel function:

In the formula, k is the Gaussian kernel radius, and σ is the standard deviation. When the convolution window slides, in view of the fact that the Gaussian kernel coefficient weight is proportional to the variance, the Gaussian kernel standard deviation σ can be obtained based on the size of the variance. The calculation formula of the variance in a certain area of the image is:

In the formula, S _i , j represents the convolution window of the center point (i, j). The larger the variance D (i, j), the greater the degree of discreteness of pixel values in the S _i , j area, and the smaller the selection. The greater the weight of the Gaussian kernel coefficient generated by σ, the smaller the impact on the region. According to this characteristic, the variance D(i, j) is compared with the two-dimensional Gaussian filter function f(i, j) to obtain the function:

In the formula, since D(i, j) is a constant, R(i, j) is a function of the Gaussian kernel radius k and the standard deviation σ, that is

When R=1, the weight of the parameters in the Gaussian kernel is closest to the weight of the pixel value. At this time, the standard deviation σ there is calculated from the variance D of the pixel value in the S(x, y) area. By analogy, iterations are repeated to form an adaptive Gaussian filter. Finally, all pixels of the cuttings image are convolved to complete the Gaussian filtering process. The filtered cuttings image is shown in Figure 9, which can be well eliminated. The Gaussian noise generated in the early image processing reduces the error in finding connected areas.

Use the bwlabel function to find connected areas, select 4 connected areas for pixel labeling, and calculate the total number of pixels P ₁ in the filled area to be 184272.

S106: After obtaining the total pixels of the cuttings in the image, calculate the area of each cutting:

In the formula, P ₁ is the total pixels of the measured cuttings, P ₂ is the total pixels of the reference square, S ₀ is the area of a single reference square, and n is the number of reference squares.

After calculating the area of each cutting, estimate the returned mud concentration:

In the formula, S ₁ is the area of cuttings on each image, S is the total area of each image, and a is the screening efficiency of the vibrating screen.

If η>5%, the wellbore is judged to be clean; if η<5%, the wellbore is judged to be unclean.

By judging the concentration of returned mud and cuttings, we can determine whether the wellbore is clean, so that we can take corresponding measures to reduce losses and enable drilling to proceed smoothly.

In this embodiment, the advantage is to make full use of the high efficiency and high precision of machine vision to collect images of cuttings on the screen surface of the vibrating screen, calculate the content of the returned mud and cuttings, and simultaneously calculate the returned mud and cuttings. concentration to determine whether the wellbore is clean. It is a non-contact measurement method that is safer and more reliable than the mass flow meter method of solid phase measurement; it is faster than the cuttings weighing method and can monitor whether the wellbore is clean in real time.

The above are only embodiments of the present invention, and common knowledge such as well-known specific structures and/or characteristics in the solutions will not be described in detail here. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the structure of the present invention. These should also be regarded as the protection scope of the present invention and will not affect the implementation of the present invention. effectiveness and patented practicality. The scope of protection claimed in this application shall be based on the content of the claims, and the specific implementation modes and other records in the description may be used to interpret the content of the claims.

Claims (5)

1. A method for monitoring wellbore cleaning in real time based on machine vision, which is characterized in that: comprises the steps of,

s101: collecting rock debris images of the screening surface of the vibrating screen;

s102: image contrast enhancement processing is carried out on the acquired image;

s103: carrying out graying and binarization on the image subjected to contrast enhancement treatment;

s104: carrying out rock debris edge detection and morphological treatment on the image subjected to the graying and binarization treatment;

s105: carrying out smoothing filter processing and pixel marking on the image after rock debris edge detection and morphological processing;

s106: calculating the area of rock debris in the image, calculating the concentration of returned mud rock debris, and judging whether the borehole is clean or not; after obtaining the total pixel points of the rock fragments in the image, calculating the area of each rock fragment:

wherein P is ₁ P is the total pixel of the rock debris to be measured ₂ To reference square total pixels, S ₀ The area of the reference square is single, and n is the number of the reference squares;

after calculating the area of each rock debris, estimating the concentration of returned mud:

wherein S is ₁ The rock debris area on each image is S, the total area of each image is S, and a is the screening efficiency of the vibrating screen;

if eta is more than 5%, judging that the borehole is in a clean state; if eta <5%, the borehole is judged to be unclean.

2. The machine vision based real time monitoring wellbore cleaning method of claim 1, wherein: in S101, the time interval of rock debris image acquisition is T/2, wherein T is the period of the conveyor belt.

3. The machine vision based real time monitoring wellbore cleaning method of claim 2, wherein: in S102, the gamma value is set to 0.2 in the process of enhancing the contrast of the rock debris image, so that the brightness is enhanced.

4. The machine vision based real time monitoring wellbore cleaning method of claim 3, wherein: in S103, the threshold value is selected to be 0.4 during the binarization processing, so as to reduce the image distortion.

5. The machine vision based real time monitoring wellbore cleaning method of claim 4, wherein: in S104, after the edge of the rock debris is detected, gap filling is performed by using a strel function, wherein a square structural element is adopted, and the width is set to be 3 pixels.