Measurement-while-drilling image acquisition and processing system and method

Abstract

The invention relates to a measurement-while-drilling image acquisition and processing technology, and discloses a measurement-while-drilling image acquisition and processing system and a measurement-while-drilling image acquisition and processing method, which solve the problems of insufficient image definition and inaccurate image recognition in a complex drilling environment and meet the real-time, continuous and high-precision lithology recognition requirements. According to the system, the image acquisition module carries out real-time shooting on the drilling process by carrying a high-speed camera on a drill bit, and real-time image data is obtained. And the adaptive parameter adjustment module adaptively adjusts various parameters required by the subsequent image processing module according to the noise intensity of the shot image data, and the parameters comprise a noise reduction parameter, a defogging parameter and an image enhancement related parameter. And finally, the image processing module carries out noise reduction, defogging and enhancement operation on the image data in sequence according to the adjusted parameters, and finally a clear rock quality image which can be directly used for lithology identification is obtained.

Description

Technical Field

[0001] The present invention relates to a measurement while drilling image acquisition and processing technology, and in particular to a measurement while drilling image acquisition and processing system and method. Background Art

[0002] Currently, in the field of geological exploration, borehole television (Borehole TV) is widely used for lithologic identification during drilling operations. Borehole TV employs a camera deployed within the borehole to capture images of the borehole wall and, using image analysis techniques, identifies and analyzes geological features such as lithology, structure, and fractures. This technology, typically activated after drilling is completed, provides visual images for lithologic identification and, under certain conditions, can assist geologists in their analysis and judgment. The image data collected by the borehole TV system provides geologists with information on lithologic changes during drilling operations. The technology is also used to detect underground structures and fractures, providing support for lithologic identification, particularly in ideal geological environments.

[0003] Although in-hole television technology has certain application value, it has many significant limitations in actual drilling operations: First, in-hole television can only be activated after drilling is completed, and cannot provide real-time monitoring during the drilling process. This means that it is impossible to obtain timely lithologic information during the drilling process, which affects the geologists' ability to make timely decisions during the drilling process.

[0004] Secondly, the complexity of the borehole environment during drilling significantly limits the effectiveness of in-hole television. Due to the often harsh conditions of low light, flying debris, and airflow disturbances during drilling, these factors severely affect image quality, resulting in blurred images and an inability to provide clear and accurate lithologic information. Interference from debris and airflow also increases image noise, further affecting the accuracy of lithologic identification.

[0005] Therefore, the existing in-hole television technology cannot effectively solve the problems of insufficient image clarity and inaccurate image recognition in complex drilling environments, resulting in the inability to meet the needs of real-time, continuous, and high-precision lithology identification, which in turn affects the safety and efficiency of drilling operations. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to propose a measurement while drilling image acquisition and processing system and method to solve the problems of insufficient image clarity and inaccurate image recognition in complex drilling environments, and to meet the needs of real-time, continuous and high-precision lithology identification.

[0007] The technical solution adopted by the present invention to solve the above technical problems is: In one aspect, the present invention provides a measurement while drilling image acquisition and processing system, comprising: Image acquisition module, used to collect real-time images inside the borehole during drilling; An adaptive parameter adjustment module, used to adaptively adjust relevant parameters in the image processing module according to the noise intensity of the acquired borehole image; The image processing module is used to perform noise reduction, defogging and enhancement processing on the collected borehole images in sequence.

[0008] Furthermore, the image acquisition module adopts a high-speed camera and is set on the drill bit.

[0009] Furthermore, the image processing module includes: an image noise reduction unit, an image defogging unit and an image enhancement unit; The image denoising unit is configured to remove coarse-grained noise from the image by performing global denoising and local denoising on the image in sequence; The image defogging unit is used to remove fine-grained noise in the image through an image defogging algorithm; The image enhancement unit is used to optimize the brightness, contrast and image edge of the image in sequence.

[0010] Furthermore, the image denoising unit performs global denoising on the image in the following manner: First, the image is mapped to the frequency space through the Discrete Cosine Transform (DCT), and then the spectrum of the image is low-pass filtered. Finally, the filtered image is subjected to the Inverse Discrete Cosine Transform (IDCT), and the spectrum is mapped back to the pixel space domain to obtain a globally denoised image. The image denoising unit performs local denoising on the image, including: performing local denoising on the image after global denoising by using Gaussian filtering; The image defogging algorithm adopted by the image defogging unit is a defogging algorithm based on dark channel prior (DCP); The image enhancement unit adopts a histogram equalization method when optimizing the brightness of the image, adopts a linear contrast adjustment method when optimizing the contrast of the image, and adopts an adaptive edge enhancement method based on a local mean when optimizing the edge of the image.

[0011] Furthermore, the relevant parameters adaptively adjusted by the adaptive parameter adjustment module include: the side length of the low-pass filter mask and the side length of the Gaussian kernel in the image denoising unit, the atmospheric illumination value in the image defogging unit, and the contrast gain factor and edge enhancement factor in the image enhancement unit.

[0012] On the other hand, the present invention also provides a method for acquiring and processing measurement while drilling images, comprising the following steps: S1. Real-time acquisition of borehole images during drilling; S2. Adaptively adjust relevant parameters in the image processing algorithm based on the noise intensity of the acquired borehole image; S3. Perform noise reduction, defogging and enhancement processing on the collected borehole images in sequence.

[0013] Furthermore, in step S3, the noise reduction processing includes: removing coarse-grained noise in the image by performing global noise reduction and local noise reduction on the image in sequence; the dehazing processing includes: removing fine-grained noise in the image by using an image dehazing algorithm; and the enhancement processing includes: optimizing the brightness, contrast and image edges of the image in sequence.

[0014] Furthermore, the method sequentially performs global denoising and local denoising on the image, including: first mapping the image into a frequency space by discrete cosine transform, then low-pass filtering the spectrum of the image, and performing an inverse two-dimensional discrete cosine transform on the filtered image, mapping the spectrum back to the pixel space domain to obtain a globally denoised image; finally, performing local denoising on the globally denoised image by using Gaussian filtering; The image defogging algorithm is a defogging algorithm based on dark channel prior; The step of sequentially optimizing the brightness, contrast, and edge of an image includes: The histogram equalization method is used to optimize the brightness of the image, the linear contrast adjustment method is used to optimize the contrast of the image, and the adaptive edge enhancement method based on local mean is used to optimize the edge of the image.

[0015] Furthermore, in step S2, the noise intensity of the acquired borehole image is estimated based on the standard deviation of the original image using the following formula: ; in, represents the noise intensity of the image inside the borehole; represents the standard deviation of the image within the borehole; Represents the standard deviation of natural clean images, using the standard deviation of the ImageNet dataset.

[0016] Furthermore, in step S2, the adaptive adjustment of relevant parameters in the image processing algorithm includes: Adaptively adjust the side length of the low-pass filter mask and the side length of the Gaussian kernel in the noise reduction process: ; ; in, represents the side length of the low-pass filter mask, represents the side length of the frequency spectrum, represents the noise intensity of the image inside the borehole; represents the side length of the Gaussian kernel, Indicates the side length of the image inside the borehole; Adaptive adjustment of atmospheric illumination values ​​in image dehazing: ; in, represents the atmospheric illumination value, Represents the pixel value of the image inside the borehole; Adaptively adjust the contrast gain factor and edge enhancement factor in image enhancement processing: ; ; in, represents the contrast gain factor, represents the edge enhancement factor.

[0017] The beneficial effects of the present invention are: (1) Realizing real-time continuous identification while drilling: The present invention uses a high-speed camera mounted on the drill bit and combined with an image processing algorithm. In particular, it can adaptively adjust the relevant parameters in the image processing algorithm according to the noise intensity of the real-time collected image, thereby dynamically optimizing the image processing method according to changes in the environment, thereby ensuring real-time collection and immediate continuous processing of lithologic image data during the drilling process, avoiding waiting time and post-cleaning processing time, thereby shortening the drilling operation cycle and reducing drilling costs. At the same time, real-time data feedback can help geologists adjust the drilling process and operation plan in a timely manner, improving the efficiency and safety of drilling operations.

[0018] (2) High-precision recognition adapted to harsh in-hole environments: In the image processing process, the present invention adopts a series of processing technologies such as noise reduction, defogging and image enhancement, which can effectively reduce the noise interference caused by complex environments, improve the brightness, contrast and clarity of images, and ensure the usability of images in harsh environments, thereby improving the accuracy and reliability of lithology analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a principle framework diagram of the measurement while drilling image acquisition and processing in the present invention. DETAILED DESCRIPTION

[0020] The present invention aims to provide a measurement-while-drilling (MWD) image acquisition and processing system and method to address the issues of insufficient image clarity and inaccurate image recognition in complex drilling environments, meeting the requirements for real-time, continuous, and high-precision lithologic identification. Its core concept is to install a high-speed camera at the drill bit to capture in-hole image data in real time during drilling. Combined with a series of advanced image noise reduction, dehazing, and image enhancement technologies, the system automatically optimizes image brightness, contrast, and clarity in harsh environments such as low light and high rock debris, thereby improving image quality. Furthermore, adaptively adjusting relevant parameters in the image processing algorithm based on the noise intensity of the real-time acquired images allows for dynamic optimization of image processing methods based on environmental changes, ensuring high-quality and efficient image processing during drilling, adapting to varying noise levels and environmental conditions, while also ensuring good continuity in image processing. Based on the above, the present invention can provide clear rock images for "real-time," "continuous," and "high-precision" lithologic identification.

[0021] In specific implementation, the principles of the measurement while drilling image acquisition and processing in the present invention are as follows: Figure 1 As shown, first, the image acquisition module captures the drilling process in real time by mounting a high-speed camera on the drill bit to obtain real-time image data. Then, the adaptive parameter adjustment module adaptively adjusts various parameters required by the subsequent image processing module according to the noise intensity of the captured image data, including noise reduction parameters, defogging parameters, and image enhancement-related parameters. Finally, the image processing module performs noise reduction, defogging, and enhancement operations on the image data in sequence according to the adjusted parameters, and finally obtains a clear rock image that can be directly used for lithologic identification. The present invention solves the problem of "inability to identify while drilling" in the prior art through the coordinated cooperation between the three modules of image acquisition, adaptive parameter adjustment, and image processing, and ensures real-time feedback of lithologic information. This enables geologists to understand lithologic changes in a timely manner during the drilling process, make quick decisions, and avoid low drilling efficiency or safety hazards caused by inaccurate geological judgments.

[0022] Example: This embodiment provides a measurement while drilling image acquisition and processing system, which includes an image acquisition module for acquiring borehole images in real time during drilling; an adaptive parameter adjustment module for adaptively adjusting relevant parameters in the image processing module based on the noise intensity of the acquired borehole images; and an image processing module for sequentially performing noise reduction, dehazing, and enhancement processing on the acquired borehole images.

[0023] The specific implementation and functions of each module are described in detail below.

[0024] 1. Image acquisition module: The image acquisition module mainly involves the deployment of high-speed cameras in the borehole and the transmission of real-time image signals. In this embodiment, the RGeo-eye high-speed cable drilling camera system is used.

[0025] The RGeo-eye high-speed wireline borehole camera system is a compact, full-color downhole observation camera with high transmission rates. It operates over 4-core or coaxial cable and is fully compatible with existing Robertson Geo winches and surface systems. Integrating the RGeo-fast module, the system achieves a 1Mb / s communication rate and can capture high-resolution video from open or cased holes, filled with air or water, downhole (up to 3,000 meters), while simultaneously providing a real-time view of surface operations. The camera features autofocus, an internal LED array on the front for adjustable illumination, and SVGA (800 x 600) viewing resolution at 25 frames per second. It also supports a "snapshot" mode for UXGA high-resolution still images (1600 x 1200), allowing for screenshots and text editing during viewing / recording. The acquisition software package allows the operator to control a wide range of features, including video resolution, frame rate, shutter speed, illumination intensity, and white balance. Four user-definable presets can be used to store optimal settings for different wellbore conditions. Video is recorded in AVI format and converted to MP4 for storage and forwarding.

[0026] 2. Image processing module: The image processing module processes the raw rock images captured by the image acquisition module's high-speed camera, removing interference from rock debris, airflow, and other factors while highlighting valuable objects within the image. The image processing module sequentially performs noise reduction, dehazing, and enhancement. Image noise reduction and dehazing, respectively, remove coarse and fine noise from the original image, eliminating the effects of factors such as rock debris, airflow, and water mist within the borehole. Image enhancement improves image clarity and texture through brightness and contrast adjustments and blur compensation (edge ​​enhancement), highlighting valuable objects within the original image.

[0027] 1. Image noise reduction unit: In the noise reduction unit, this embodiment adopts the idea of ​​"global first, then local" to perform global and local noise reduction on the image in sequence, ensuring that coarse-grained noise in the original image is effectively removed.

[0028] During the global denoising process, this embodiment first maps the image into frequency space using a discrete cosine transform (DCT). The image's spectrogram is then low-pass filtered to remove global high-frequency noise. Finally, an inverse two-dimensional discrete cosine transform is performed on the filtered image, and the spectrogram is remapped back to the pixel space domain to obtain the globally denoised image.

[0029] Assume that the image to be processed is Resolution, the mathematical expression is as follows: (1); (2); (3); in, and Represent the pixel domain image and the corresponding frequency domain spectrum respectively; Represents the result after noise reduction; and Represents the pixel spatial position and the spectrum graph in Different frequencies in different directions; is the low-pass filter mask, represents Hadamard Product; is a constant coefficient used for normalization, specifically: (4); After global noise reduction, this embodiment uses Gaussian filtering to perform local noise reduction on the image, further optimizing and removing local noise. Its mathematical expression can be: (5); in, Express the results after processing; " is numerical multiplication; represents the Gaussian kernel The side length of Gaussian kernel The mathematical expression can be written as: (6); After global and local double noise reduction processing, the coarse-grained noise in the original image will be eliminated.

[0030] There are two parameters that need to be given in the image denoising unit: low-pass filter mask The side length and Gaussian kernel The side length These two parameters will be adaptively adjusted for images with different noise intensities in the adaptive parameter adjustment module.

[0031] 2. Image defogging unit: After removing coarse-grained noise from the image, the next step is to remove fine-grained noise. Fine-grained noise can be caused by edge effects of coarse-grained noise or by haze caused by rock fragments or water mist within the hole. Due to its small and dense nature, this embodiment uses an image dehazing algorithm to address fine-grained noise.

[0032] The Dark Channel Prior (DCP) dehazing algorithm is an image dehazing method based on the statistical properties of natural images. It exploits a simple assumption: in a natural image without haze, the dark channel values ​​are very low in at least some areas, while the presence of haze significantly increases these values. Using this prior, DCP can effectively estimate the image's transmittance and perform dehazing. The core idea of ​​the dark channel prior is that in a natural image without haze, the pixel values ​​of at least one color channel (red, green, or blue) are always very low in certain areas. Generally speaking, background portions of an image (such as the sky and trees) have very low pixel values, especially in areas with richer colors.

[0033] For a given image , the dark channel value is calculated by the following formula: (7); in, Is the image at position Color channels (red, green, or blue) pixel value; Indicates The local neighborhood centered on is generally implemented in the form of a local window (usually or (a small window appears). Indicates that the image is The minimum value of the RGB channels within the neighborhood is the dark channel value of the image. This "dark channel" value reflects the darkest pixel value in the image and is an important prior in the dehazing algorithm.

[0034] In order to achieve image dehazing, it is necessary to mathematically model the haze image. The haze image can be mathematically expressed as: (8); in, and Represent foggy images and fog-free images respectively; It's a pixel The transmittance at ,indicates the degree of influence of haze on the pixel; is the global atmospheric illumination value (often the maximum brightness value in the image, usually assumed to be the brightness of the sky). Represents the distribution of haze intensity, usually with values ​​between 1 means no haze, and 0 means complete obstruction.

[0035] Transmittance in modeling haze images Still an unknown quantity, we need to estimate the transmittance here. The estimation formula is: (9); in, is a constant (typically set to 0.95) that represents a tuning parameter controlling the transmittance estimate. This equation implies that transmittance is directly related to the ratio of the local minimum dark channel value to the global atmospheric illumination value. Low dark channel values ​​(i.e., areas free of haze) correspond to higher transmittance, while high dark channel values ​​(i.e., areas with greater haze influence) correspond to lower transmittance.

[0036] Finally, after estimating the transmittance and global atmospheric lighting values (generally taking the maximum brightness value in the image), the haze image model can be used to restore the real image : (10); This formula restores the foggy image based on the transmittance and atmospheric illumination value to obtain the defogged image. .

[0037] After the above steps, the defogging unit can perform effective defogging based on the dark channel prior.

[0038] In this process, there are two parameters that need to be adjusted: transmittance and atmospheric illumination values .

[0039] Among them, the transmittance It can be estimated according to formula (9), and the atmospheric illumination value Adaptive adjustment needs to be made in the adaptive parameter adjustment module for different haze intensities.

[0040] 3. Image enhancement unit: The drilling process often faces harsh conditions such as low light, flying rock debris, and airflow disturbances, which severely impact image quality, resulting in blurred images and a lack of clear and accurate lithologic information. The image noise reduction unit and defogging unit perform coarse and fine-grained noise reduction and defogging on the raw images captured by the high-speed camera, respectively. The image enhancement unit enhances image light perception, color, and edges, highlighting valuable objects in the original image and improving the accuracy of lithologic identification. The image enhancement unit sequentially optimizes image brightness, contrast, and edges, effectively addressing the challenges of low light and image blur during drilling.

[0041] This embodiment uses histogram equalization (Histogram Equalization) to adjust image brightness. Histogram equalization is a common image enhancement technique that aims to enhance image contrast by adjusting the grayscale distribution of an image, thereby making image details more visible. The basic idea of ​​histogram equalization is to stretch the grayscale value range of an image to make the image's grayscale histogram more evenly distributed.

[0042] First, we need to calculate the grayscale histogram of the image. Assume that the grayscale of the input image is (For example, the grayscale range of an 8-bit image is 0 to 255.) The grayscale value of each pixel in the image is , then the grayscale histogram of the image Indicates the gray value is The frequency of pixels. Next, we need to calculate the cumulative distribution function (CDF). is the cumulative sum of the grayscale histogram, defined as: (11); in, is the grayscale value The number of pixels; is the total number of pixels in the image. Cumulative distribution function Describes grayscale values ​​less than or equal to Finally, we need to map the original image pixel grayscale value to the new grayscale level. We can use the cumulative distribution function Map the grayscale value of the original image. The new grayscale value It can be calculated according to the following formula: (12); in, is the maximum grayscale value (256 for 8-bit images); yes The minimum value in ; represents the rounding function. This formula is obtained by Linearly map to the new gray level to achieve gray level equalization.

[0043] After histogram equalization, image pixels are evenly distributed, effectively balancing the brightness and darkness of the image and providing a good pre-processing step for increasing saturation. Histogram equalization does not require manual adjustment of parameters, so the brightness adjustment part does not interact with the adaptive parameter adjustment module.

[0044] This embodiment uses linear contrast adjustment to adjust image contrast. Linear contrast adjustment is a common image processing technique that adjusts image contrast through linear mapping. Its basic concept is to amplify or reduce grayscale differences in the image by changing its grayscale distribution, thereby enhancing or weakening the image contrast.

[0045] Assume that the grayscale value of the input image is The range is , the output grayscale value of linear contrast adjustment This can be achieved through the following formula: (13); in, is the contrast gain factor, which determines the intensity of image contrast; is the brightness offset, which is used to adjust the brightness of the image so that its grayscale value remains within the valid range. Gain that controls image contrast. If , the image contrast increases and the image becomes more vivid. On the contrary, if , the contrast of the image is reduced, the image becomes smoother, and the difference between gray values ​​becomes smaller. Adjusts the brightness of an image. Typically, Used to limit the grayscale value of the adjusted image to For example, when the minimum grayscale value of the image is 0 and the maximum grayscale value is 255, you can set To make the result after linear transformation within this range.

[0046] Since the overall brightness of the image has been adjusted when adjusting the contrast in this embodiment, the brightness offset It can be dynamically adjusted according to the existing brightness, so as not to destroy the brightness structure adjusted in the previous step. The specific mathematical expression is as follows: (14); in, is the average gray value of the original image; is the target brightness. Indicates the average brightness range of the image after processing. This is a hyperparameter that needs to be set before the algorithm is executed. Considering the low light, rock cuttings, airflow and other factors in the drilling process, this embodiment defaults to the general target brightness. It is set to 164, and a slightly higher average brightness level is beneficial to the accuracy of the final lithology identification.

[0047] The parameters that need to be adjusted in contrast adjustment are: contrast gain factor Target brightness .

[0048] Due to the target brightness With default setting value, there is no need to adjust the adaptive parameter adjustment module. Here there is only contrast gain factor It needs to be determined through adaptive adjustment.

[0049] This embodiment uses local mean-based adaptive edge enhancement (LME) for image edge enhancement. Adaptive edge enhancement is an image processing method that adaptively adjusts the edge enhancement strength based on local image features, such as local brightness, contrast, or texture. Unlike traditional global edge enhancement methods, this adaptive method can adjust the enhancement effect based on the image characteristics of different regions, enhancing contrast in edge areas while maintaining smoothness in flat areas, thereby achieving a more natural and refined image enhancement effect.

[0050] Specifically, the image edge enhancement process can be described by the following mathematical expression: (15); in, and Represent the original image pixel grayscale value and the pixel grayscale value after edge enhancement respectively; is an enhancement factor used to control the strength of the enhancement; For the original image This method provides greater enhancement in the edge areas of the image (areas with high mean difference) and less enhancement in flat areas (areas with low mean difference), thus avoiding over-enhancement of noise.

[0051] Since the only parameter that needs to be adjusted for edge enhancement is the edge enhancement factor Therefore, there is only the edge enhancement factor It needs to be determined by the adaptive adjustment module.

[0052] 3. Adaptive Parameter Adjustment Module To overcome the challenges of real-time and discontinuous detection in drilling (MWD) systems, relying solely on high-speed cameras and high-speed communications is not enough. The system also needs to adaptively adjust the specific parameters of each imaging algorithm to accommodate varying noise and light intensities to achieve effective real-time and continuous MWD.

[0053] Based on the above, the image processing algorithm parameters that need to be adaptively adjusted are as follows: (1) Low-pass filter mask The side length ; (2) Gaussian kernel The side length ; (3) Atmospheric illumination value ; (4) Contrast gain factor ; (5) Edge enhancement factor ; in, and These are the parameters required for global and local denoising respectively; are the parameters required by the image dehazing algorithm; and These are the parameters for contrast adjustment and edge enhancement.

[0054] Adaptive parameter adjustment requires adjusting subsequent parameters based on the noise intensity of the input image. Since clean images are unavailable, this embodiment adaptively adjusts various parameters of the image processing algorithm by estimating the noise intensity, given the original image. Considering that clean images have low noise levels and low standard deviations, while noisy images have high standard deviations and high noise levels, this embodiment estimates the noise intensity by calculating the standard deviation of the original image.

[0055] for The standard deviation of the image resolution is: (16); (17); in, and Represent the image mean and variance respectively. Since the pixel value range of the original image is In order to facilitate subsequent calculations, it is necessary to perform Min-Max Scaling on the pixels before calculating the mean and variance, so that their value range remains within The specific expression is: (18); in, and Respectively represent the pixel grayscale values ​​of the original image before and after normalization; and Represent the minimum and maximum grayscale values ​​of image pixels respectively.

[0056] The ImageNet dataset is one of the most important and commonly used image databases in the field of computer vision. It is widely used in research and algorithm development for tasks such as image classification, object detection, and image segmentation. As a representative natural image dataset, this example uses the standard deviation of the ImageNet dataset as the standard deviation of clean images as a benchmark reference.

[0057] The standard deviation of the RGB three-channel images in the ImageNet dataset is In order to simplify the calculation, this embodiment sets the standard deviation of the natural image to Therefore, the noise intensity calculated in this embodiment is: (19); in, and Represents the standard deviation of noisy images and natural clean images, and the range of both is From the above formula, we can know the noise intensity The value range of . A value close to 0 indicates that the image noise is very small and close to the natural image, while a value close to 1 indicates that the image noise is very large and far from the natural image.

[0058] After obtaining the estimated value of the noise intensity of the original image After that, the various parameters in the image processing algorithm can be adaptively set according to their values. The specific adjustments are as follows: (20); (twenty one); (twenty two); (twenty three); (twenty four); in, Indicates the side length of the frequency spectrum, the larger will make Smaller means more high-frequency noise needs to be filtered out; Represents the side length of the noisy image, the larger will make Larger means using a wider range of pixels for noise reduction, resulting in stronger noise reduction. Represents the pixel value of the noisy image and the global atmospheric illumination value in dark channel defogging The maximum brightness value in the image is often taken, usually assumed to be the brightness of the sky; the contrast gain factor When the contrast is increased, since larger noise will make the image darker, a larger For larger contrast gains ; Since larger noise will blur the edge of the picture, larger For larger edge enhancement factors .

[0059] Based on the above-mentioned measurement while drilling image acquisition and processing system, the measurement while drilling image acquisition and processing method provided in this embodiment includes the following steps: S1. Real-time acquisition of borehole images during drilling; S2. Adaptively adjust relevant parameters in the image processing algorithm based on the noise intensity of the acquired borehole image; S3. Perform noise reduction, defogging and enhancement processing on the collected borehole images in sequence.

[0060] It is understandable that since the functions of the various functional modules of the measurement while drilling image acquisition and processing system in this embodiment correspond to the step descriptions of the measurement while drilling image acquisition and processing method in this embodiment, based on the above detailed description of the functional implementation of each functional module of the system, the specific implementation of each step of the measurement while drilling image acquisition and processing method will not be repeated here.

[0061] Finally, it should be noted that the above embodiments are merely preferred implementations and are not intended to limit the present invention. It should be noted that those skilled in the art will be able to make modifications, equivalent substitutions, and improvements without departing from the spirit and scope of the present invention and the claims, all of which should be included within the scope of protection of the present invention.

Claims

1. A measurement while drilling image acquisition and processing system, characterized in that: include: Image acquisition module, used to collect real-time images inside the borehole during drilling; An adaptive parameter adjustment module, used to adaptively adjust relevant parameters in the image processing module according to the noise intensity of the acquired borehole image; The image processing module is used to perform noise reduction, defogging and enhancement processing on the collected borehole images in sequence.

2. The measurement while drilling image acquisition and processing system according to claim 1, characterized in that: The image acquisition module adopts a high-speed camera and is arranged on the drill bit.

3. The measurement while drilling image acquisition and processing system according to claim 1, characterized in that: The image processing module includes: an image noise reduction unit, an image defogging unit and an image enhancement unit; The image denoising unit is configured to remove coarse-grained noise from the image by sequentially performing global denoising and local denoising on the image; The image defogging unit is used to remove fine-grained noise in the image through an image defogging algorithm; The image enhancement unit is used to optimize the brightness, contrast and image edge of the image in sequence.

4. The measurement while drilling image acquisition and processing system according to claim 3, characterized in that: The image denoising unit performs global denoising on the image in the following manner: First, the image is mapped to the frequency space through discrete cosine transform, then the spectrum of the image is low-pass filtered, and finally the filtered image is inversely transformed into a two-dimensional discrete cosine transform, and the spectrum is mapped back to the pixel space domain to obtain a globally denoised image. The image denoising unit performs local denoising on the image, including: performing local denoising on the image after global denoising by using Gaussian filtering; The image defogging algorithm adopted by the image defogging unit is a defogging algorithm based on dark channel prior; The image enhancement unit adopts a histogram equalization method when optimizing the brightness of the image, adopts a linear contrast adjustment method when optimizing the contrast of the image, and adopts an adaptive edge enhancement method based on a local mean when optimizing the edge of the image.

5. The measurement while drilling image acquisition and processing system according to claim 4, characterized in that: The relevant parameters adaptively adjusted by the adaptive parameter adjustment module include: the side length of the low-pass filter mask and the side length of the Gaussian kernel in the image denoising unit, the atmospheric illumination value in the image defogging unit, and the contrast gain factor and edge enhancement factor in the image enhancement unit.

6. A method for acquiring and processing measurement while drilling images, characterized in that: The following steps are involved: S1. Real-time acquisition of borehole images during drilling; S2. Adaptively adjust relevant parameters in the image processing algorithm based on the noise intensity of the acquired borehole image; S3. Perform noise reduction, defogging and enhancement processing on the collected borehole images in sequence.

7. The method for acquiring and processing measurement while drilling images according to claim 6, wherein: In step S3, the noise reduction processing includes: removing coarse-grained noise in the image by performing global noise reduction and local noise reduction on the image in sequence; the dehazing processing includes: removing fine-grained noise in the image by using an image dehazing algorithm; and the enhancement processing includes: optimizing the brightness, contrast and image edges of the image in sequence.

8. The method for acquiring and processing measurement while drilling images according to claim 7, wherein: The method sequentially performs global denoising and local denoising on the image, including: first mapping the image into a frequency space by discrete cosine transform, then low-pass filtering the spectrum of the image, and performing an inverse two-dimensional discrete cosine transform on the filtered image, mapping the spectrum back to the pixel space domain to obtain a globally denoised image; finally, performing local denoising on the globally denoised image by Gaussian filtering; The image defogging algorithm is a defogging algorithm based on dark channel prior; The step of sequentially optimizing the brightness, contrast, and edge of an image includes: The histogram equalization method is used to optimize the brightness of the image, the linear contrast adjustment method is used to optimize the contrast of the image, and the adaptive edge enhancement method based on local mean is used to optimize the edge of the image.

9. The method for acquiring and processing measurement while drilling images according to claim 7, wherein: In step S2, the noise intensity of the acquired borehole image is estimated based on the standard deviation of the original image using the following formula: in, represents the noise intensity of the image inside the borehole; represents the standard deviation of the image within the borehole; Represents the standard deviation of natural clean images, using the standard deviation of the ImageNet dataset.

10. The method for acquiring and processing measurement while drilling images according to claim 9, wherein: In step S2, the adaptive adjustment of relevant parameters in the image processing algorithm includes: Adaptively adjust the side length of the low-pass filter mask and the side length of the Gaussian kernel in the noise reduction process: ； ； in, represents the side length of the low-pass filter mask, represents the side length of the frequency spectrum, represents the noise intensity of the image inside the borehole; represents the side length of the Gaussian kernel, Indicates the side length of the image inside the borehole; Adaptive adjustment of atmospheric illumination values ​​in image dehazing: ； in, represents the atmospheric illumination value, Represents the pixel value of the image inside the borehole; Adaptively adjust the contrast gain factor and edge enhancement factor in image enhancement processing: ； ； in, represents the contrast gain factor, represents the edge enhancement factor.

Images