CN110689512B - Method for quickly splicing and fusing annular images of panoramic video in hole into image - Google Patents
Method for quickly splicing and fusing annular images of panoramic video in hole into image Download PDFInfo
- Publication number
- CN110689512B CN110689512B CN201910903069.8A CN201910903069A CN110689512B CN 110689512 B CN110689512 B CN 110689512B CN 201910903069 A CN201910903069 A CN 201910903069A CN 110689512 B CN110689512 B CN 110689512B
- Authority
- CN
- China
- Prior art keywords
- image
- narrow
- band
- offset
- matching
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 238000007526 fusion splicing Methods 0.000 claims abstract description 14
- 238000005553 drilling Methods 0.000 claims description 31
- 238000005259 measurement Methods 0.000 claims description 28
- 230000004927 fusion Effects 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 11
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000003247 decreasing effect Effects 0.000 claims 1
- 238000012545 processing Methods 0.000 abstract description 12
- 230000008569 process Effects 0.000 abstract description 10
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 31
- 230000000694 effects Effects 0.000 description 25
- 239000011435 rock Substances 0.000 description 13
- 239000004575 stone Substances 0.000 description 4
- 238000011835 investigation Methods 0.000 description 3
- 230000007704 transition Effects 0.000 description 3
- 238000013519 translation Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 239000013049 sediment Substances 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/05—Geographic models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Remote Sensing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Graphics (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for quickly splicing and fusing annular images of an intra-hole panoramic video into an image, which is used for obtaining a frame narrow-band image and obtaining azimuth data and depth data; detecting and matching features; determining a width actual offset and a height actual offset of the narrow banded image; carrying out fusion splicing on the narrow-band images; the invention does not depend on an auxiliary physical compass or an electronic compass and a depth encoder; the method is simple and easy to operate, can realize intelligent and automatic analysis processing in the process of splicing the map, greatly reduces the burden of workers, improves the working efficiency, and ensures that the definition and the accuracy of the spliced and fused image are far higher than those of the original scanning line map method.
Description
Technical Field
The invention belongs to the cross field of geotechnical engineering geological exploration and image information processing, and particularly relates to a method for quickly splicing and fusing annular images of panoramic videos in holes into a map.
Background
In the fields of geotechnical engineering, geological engineering, hydroelectric engineering, oil development, geological disaster prevention and control engineering and the like, the stability characteristics of a rock mass structure are required to be known frequently, and structural planes such as joints, faults, weak planes, bedding planes and the like existing in the rock mass are important factors for determining the stability of practical engineering. The digital panoramic drilling shooting system can obtain high-precision panoramic video images of the internal structural surface of the drilled hole through an in-hole shooting technology, and the panoramic video images accurately record the characteristics of the structural surface of the rock body in the hole. In engineering practice, accurate identification of the structural surface of the intra-hole panoramic video image and rapid extraction of the characteristic parameters of the structural surface have important practical engineering significance for exploration, engineering design, construction and the like.
At present, the analysis and processing of the panoramic borehole image in the hole is still basically in the scanning line image method in the traditional sense, i.e. the scanning lines of each frame of video image are continuously generated completely depending on the azimuth angle of the compass and the depth information of the encoder, and then the scanning lines generated by each frame of image are sequentially piled up to form the corresponding borehole image. The method has the following defects: (1) the scanning line of the traditional method is limited in data volume carried by the scanning line, and the field actual investigation and recording of the video image are difficult to guarantee to be carried out at a very slow pay-off speed at a constant speed, so that the video image is fast and slow in time, the scanning line cannot record more image line number data when the pay-off is too fast, and further the finally obtained drilling image is seriously deformed, time-lapse and poor in image quality; (2) the method for mapping by scanning lines completely depends on data of a compass and an encoder, and in the actual situation, the compass and the encoder have certain physical inertia and are easy to make mistakes, and in the situation, the method for mapping by scanning lines cannot modify the compass and find the correct position and analyze; (3) after the software of the scanning line method is realized, self-adaption and intelligent processing cannot be achieved, and manual processing is seriously depended on, so that for video image data of a plurality of holes or deep holes, special processing, time and labor consumption are needed for days or even weeks, and intelligent and automatic analysis processing of video images in the holes is urgently needed.
Therefore, aiming at the problems, the invention develops a novel method for quickly splicing and fusing the video images in the hole to form a picture, aims to solve the problems of intelligently matching the image data and automatically splicing and fusing the image data of the panoramic video images in the hole, solves the problem of completely depending on a compass or an electronic compass, realizes the intelligent splicing and fusing and real-time quick drawing of the panoramic video images in the hole, reduces the burden of engineering researchers, improves the working efficiency, provides great convenience for the actual drilling exploration process, and greatly improves the real-time property and the high efficiency of the camera exploration in the hole of the deep rock mass structure.
Disclosure of Invention
The invention aims to provide a method for quickly splicing and fusing annular images of panoramic videos in holes into a picture aiming at the defects in the prior art. The method has the advantages of high automation degree and high reliability, splicing and fusing two adjacent frames of images by utilizing the inherent characteristics of the narrow-band images of the drilling images of the in-hole panoramic video, and the like, thereby realizing continuous splicing and fusing of thousands of and tens of thousands of frames of narrow-band images in the in-hole panoramic video images and finally quickly forming the fused and spliced images required by the engineering. The invention greatly improves the working efficiency and accuracy of the drilling exploration and the image data analysis process.
In order to solve the technical problems, the invention adopts the following technical scheme:
a method for quickly splicing and fusing annular images of panoramic videos in holes into an image comprises the following steps:
step 1, obtaining each frame of drilling image obtained through an in-hole panoramic video to obtain each corresponding frame of narrow-band-shaped image, and obtaining Azimuth angle data Azimuth of a compass or an electronic compass in each frame of drilling image and Depth data Depth of a Depth encoder;
step 2, feature detection and matching, which specifically comprises the following steps:
step 2.1, sequentially taking three adjacent frame narrow strip-shaped images, which are respectively defined as a first narrow strip-shaped image Img1, a second narrow strip-shaped image Img2 and a third narrow strip-shaped image Img 3; the first difference image ImgA is a difference value between the second narrowband image Img2 and the first narrowband image Img1, that is, ImgA is Img2 to Img 1; the second difference image ImgB is a difference value of the third narrow band-shaped image Img3 and the second narrow band-shaped image Img2,
step 2.2, the first difference image ImgA is converted into a first gray level image gray a, the second difference image ImgB is converted into a second gray level image gray b,
step 2.3, detecting a first gray image feature point and a first gray image feature description operator in the first gray image GrayA, detecting a second gray image feature point and a second gray image feature description operator in the second gray image GrayB,
step 2.4, calculating the Hamming distance from each first gray image feature point to each second gray image feature point according to the first gray image feature description operator and the second gray image feature description operator,
step 2.5, the minimum Hamming distance is an optimal matching distance, and a first gray image characteristic point and a second gray image characteristic point corresponding to the optimal matching distance are called as optimal characteristic matching point pairs and are respectively marked as Ma and Mb; the sub-minimum Hamming distance is used as a sub-optimal matching distance, and a first gray image feature point and a second gray image feature point corresponding to the sub-optimal matching distance are called as sub-optimal feature matching point pairs and are respectively marked as Ma 'and Mb';
step 2.6, taking the average value of the optimal characteristic matching point pair and the suboptimal characteristic matching point pair as an average characteristic matching point pair,
determining corresponding matched pixel points on the second narrow banded image Img2 and the third narrow banded image Img3 according to the first gray image characteristic point and the second gray image characteristic point corresponding to the optimal characteristic matching point pair or the average characteristic matching point pair,
calculating a width matching offset in the width direction of the matching pixel points of the second narrowband image Img2 with respect to the matching pixel points on the third narrowband image Img3, and a height matching offset in the height direction,
step 3, calculating Azimuth pixel measurement offset and Depth pixel measurement offset of the narrowband image according to the Azimuth data Azimuth and the Depth data Depth of the Depth encoder, determining width actual offset and height actual offset of the narrowband image,
and 4, fusing and splicing the narrow strip image of the current frame and the narrow strip image of the next frame according to the actual width offset and the actual height offset.
Step 1 as described above comprises the steps of:
identifying a center O (x, y) of the borehole image and a maximum radius Rmax and a minimum radius Rmin of a borehole wall annular image within the borehole image, wherein x and y are respectively an abscissa and an ordinate of the borehole image,
on the annular image of the hole wall of the drill hole, W pixel points are respectively collected on each circle with the radius of Rmin-Rmax, the W pixel points collected on each circle form corresponding pixel rows, a narrow-band-shaped image is generated according to each pixel row, the corresponding radius of the uppermost row to the lowermost row of the narrow-band-shaped image is from large to small, the width of the narrow-band-shaped image is W, the height of the narrow-band-shaped image is H, and H is Rmax-Rmin.
Step 3 as described above comprises the steps of:
step 3.1, obtaining Azimuth angle data Azimuth and Depth data Depth of each frame of narrow-band-shaped image, and calculating the Azimuth angle value Azimuth of the current frame of narrow-band-shaped imageiAzimuth value Azimuth of narrow banded image with next framei+1Is used as the Azimuth measurement offset delta Azimuth of the narrow-band image of the current frameiI is the number of frames of the narrow band image,
calculating Depth data Depth of current frame narrow-band imageiDepth data Depth of narrow banded image with next framei+1Is used as the Depth measurement offset delta Depth of the narrow-band image of the current framei，
Measuring the offset Δ Azimuth by the Azimuth of the narrow banded image of the current frameiAnd the Depth measurement offset Δ Depth of the current frame narrow-band imageiCalculate an azimuth pixel measurement offset and a depth pixel measurement offset,
step 3.2, if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is smaller than W/4, the actual width offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is the width matching offset; if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is larger than or equal to W/4 and smaller than W, the actual width offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is an azimuth pixel measurement offset; if the absolute value of the width matching offset of the current frame narrow-band image relative to the next frame narrow-band image in the width direction is greater than or equal to W, the width actual offset of the current frame narrow-band image relative to the next frame narrow-band image in the width direction is W,
if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is smaller than the first height pixel threshold value, the actual height offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is the height matching offset; if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is greater than or equal to a first height pixel threshold value and smaller than a second height pixel threshold value, the actual height offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is a depth pixel measurement offset; and if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is greater than or equal to the second height pixel threshold value, the height actual offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is H.
Both the Azimuth data Azimuth and the Depth data Depth of the narrowband image of each frame in step 3.1 are smoothed as described above.
Step 4 as described above comprises the steps of:
circularly translating the current frame narrow band-shaped image according to the actual width offset, correcting the offset of the current frame narrow band-shaped image and the next frame narrow band-shaped image in the width direction, performing weighted fusion on the overlapped part of the circularly translated current frame narrow band-shaped image and the next frame narrow band-shaped image in the height direction, wherein the pixel height of the overlapped part of the circularly translated current frame narrow band-shaped image and the next frame narrow band-shaped image in the height direction is the actual height offset,
and acquiring a spliced fusion spliced image of each frame of narrow band-shaped image, and dividing the fusion spliced image into a plurality of fusion spliced sub-images according to the set depth length.
And in the direction from the current frame narrow-band-shaped image to the next frame narrow-band-shaped image, the first weighting coefficients of the overlapped parts of the current frame narrow-band-shaped image and the next frame narrow-band-shaped image in the height direction after circular translation are sequentially reduced, the second weighting coefficients are sequentially increased, and the sum of the first weighting coefficient and the second weighting coefficient is 1.
A method for quickly splicing and fusing annular images of an in-hole panoramic video into an image further comprises the following steps of 5: and 4, performing image enhancement and length-width pixel size conversion on each of the multiple fusion splicing subimages formed in the step 4.
A method for quickly splicing and fusing annular images of an in-hole panoramic video into an image further comprises the following steps of 6: and marking Azimuth data Azimuth and Depth data Depth of each fusion splicing subimage, and generating a drilling histogram in four directions of east, south, west and north respectively.
Related concept definition and annotation:
(1) the structural surface in the present invention is a geological concept.
(2) The borehole image is equivalent to a borehole wall image, a panoramic hole wall image and an in-hole panoramic hole wall image.
(3) The drilling image is divided into a high-definition image and a standard-definition image; the two video images are not different in nature, except that the video image quality is high or low, the definition is different, and the video data volume is large or small, an electronic compass is adopted for azimuth angle information in 200 ten thousand high-definition videos, and a physical compass is adopted for azimuth angle information in 50 ten thousand high-definition videos.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the method does not depend on an auxiliary physical compass or an electronic compass and a depth encoder to carry out continuous splicing and fusion of the borehole images in the hole. The scanline method relying entirely on a compass and an electronic compass is abandoned.
2. The method is simple and easy to operate, can realize intelligent and automatic analysis processing in the process of splicing the pictures, greatly reduces the burden of workers and improves the working efficiency.
3. The method can avoid the borehole image with stretching deformation caused by the verticality of a serious depth encoder, a compass or an electronic compass, and has poor image quality and low definition. The invention automatically matches data according to the fixed specification of each frame of narrow-band image, realizes continuous splicing and forms an undifferentiated high-quality original image.
4. The quality of the finally formed spliced image is far higher than that of the traditional scanning line method, and an effect figure is shown in figure 1. Fig. 1(a) is a comparison diagram of an original image effect diagram of the present invention and an original image effect diagram of a scanning line method in a certain area of a drilling image formed after the same drilling video is processed by the present invention and the original scanning line method, wherein the left diagram is the original image effect diagram of the present invention, and the right diagram is the original image effect diagram of the scanning line method; fig. 1(b), fig. 1(c), and fig. 1(d) are comparison diagrams of enlarged views of the same caving stone block, the same hole, and the same rock crack in fig. 1(a), respectively, and it can be known that, after the same position is viewed and compared with the same magnification, the definition and accuracy of the spliced fused image obtained by the image matching fusion method of the present invention are far higher than those of the original scan line image method.
5. The invention has high automation degree, is convenient for realizing full-automatic treatment and programming treatment, simplifies the operation process and greatly improves the working efficiency.
Drawings
Fig. 1(a) is a comparison diagram of an original image effect diagram of the present invention and an original image effect diagram of a scanning line method in a certain area of a drilling image formed after the same drilling video is processed by the present invention and the original scanning line method, wherein the left diagram is the original image effect diagram of the present invention, and the right diagram is the original image effect diagram of the scanning line method;
fig. 1(b) is a comparison graph of enlarged views of the same collapsed stone in the original image effect graph of the present invention and the original image effect graph of the scanning line method, wherein the left graph is an enlarged view of the collapsed stone in the original image effect graph of the present invention, and the right graph is an enlarged view of the same collapsed stone in the original image effect graph of the scanning line method;
fig. 1(c) is a comparison diagram of enlarged views of the same hole in the original image effect diagram of the present invention and the original image effect diagram of the scan line method, wherein the left diagram is an enlarged view of the hole in the original image effect diagram of the present invention, and the right diagram is an enlarged view of the same hole in the original image effect diagram of the scan line method;
fig. 1(d) is a comparison diagram of enlarged views of the same rock crack in the original image of the invention and the original image of the scanning line method, wherein the left diagram is an enlarged view of the rock crack in the original image of the invention, and the right diagram is an enlarged view of the same rock crack in the original image of the scanning line method;
fig. 2 is a schematic diagram of basic information acquisition of 200-ten-thousand high-definition intra-hole panoramic video.
FIG. 3(a) is a comparison graph of the first grayscale image feature point and the second grayscale image feature point on the first grayscale image and the second grayscale image; in the figure, the upper graph is a schematic diagram of a first gray scale image characteristic point of a first gray scale image, and the lower graph is a schematic diagram of a second gray scale image characteristic point of a second gray scale image.
Fig. 3(b) is a diagram of positions and associations of the first grayscale image feature point and the second grayscale image feature point on the current frame narrowband image and the next frame narrowband image, respectively.
Fig. 4 is a schematic diagram of an optimal feature matching point pair on a current narrow-band image of a current frame and a current narrow-band image of a next frame.
FIG. 5 is a comparison graph of the effect of an image obtained by directly stitching a previous frame of narrowband image and a next frame of narrowband image and an image obtained by weighted fusion stitching; wherein, the left image is the directly spliced image, and the right image is the weighted fusion spliced image.
Fig. 6 shows 4360 narrow band images in the example to form 2 fused stitching sub-images.
Fig. 7 is a fused stitched subimage that has undergone gray scale stretching and gray scale equalization.
Fig. 8 is a borehole histogram of fused mosaic sub-images, wherein (a) and (b) are one of the fused mosaic sub-images and the corresponding borehole histogram, respectively, and (c) and (d) are the other of the fused mosaic sub-images and the corresponding borehole histogram, respectively.
Detailed Description
The present invention will be described in further detail with reference to examples for the purpose of facilitating understanding and practice of the invention by those of ordinary skill in the art, and it is to be understood that the present invention has been described in the illustrative embodiments and is not to be construed as limited thereto.
Example 1: aiming at the situation that 200 ten thousand pixels of high-definition in-hole panoramic video and an electronic compass are carried
A method for quickly splicing and fusing annular images of panoramic videos in holes into an image comprises the following specific steps on the premise that the panoramic video videos in the holes are obtained, and detailed specific embodiment explanation is carried out by taking actually-measured drilling video image data as an example.
Aiming at the conditions that 200 ten thousand high-definition intra-hole panoramic videos are carried with an electronic compass, a method for quickly splicing and fusing annular images of the intra-hole panoramic videos into a picture comprises the following main steps:
step 1, narrow-band image acquisition
For a borehole image obtained by acquiring 200-kilo-resolution intra-borehole panoramic video, firstly, a hough circle monitoring method (such as HoughCircles () function in OpenCV) is utilized to automatically identify the center O (x, y) of the borehole image and the maximum radius Rmax and the minimum radius Rmin of a borehole wall annular image in the borehole image, wherein x and y are respectively the abscissa and the ordinate of the borehole image, and identify Azimuth data Azimuth of a compass or an electronic compass and Depth data Depth of a Depth encoder in the borehole image, as shown in fig. 2. The intra-hole panoramic video in the picture is recorded by 200 high-definition drilling camera equipment and is a product developed by Chinese academy rock and soil.
In order to ensure that the narrow-band image to be generated is effective as much as possible, the effective range of the annular image of the hole wall of the drill hole is set to be H-Rmax-Rmin, and H is the height of the narrow-band image;
on the annular image of the hole wall of the drill hole, W pixel points are respectively collected on each circle which takes the center O (x, y) as the center of a circle and takes the radius Rmin-Rmax according to a set central angle clockwise (or anticlockwise). The W pixel points collected on each circle form corresponding pixel rows, narrow-band images are generated according to the pixel rows, and the corresponding radiuses from the top row to the bottom row of the narrow-band images are from large to small.
The width of the narrow band image is W, the height is H, and the height and the width of the narrow band image are pixel values.
And obtaining narrow strip-shaped images corresponding to the drilling hole wall annular images in each frame of drilling hole images of the in-hole panoramic video, wherein the drilling hole azimuth angle and the collection direction (clockwise or anticlockwise) of the pixel points in the process of obtaining each narrow strip-shaped image are the same.
Step 2, feature detection and matching data generation
Aiming at 4360 formed narrow-band images, in order to give consideration to the accuracy and timeliness of image feature detection and feature matching, the invention mainly adopts ORB features and BRIEF feature description operators to quickly detect feature points in the narrow-band images and perform feature matching.
Because narrow-band images formed by the borehole images of each frame of the in-hole panoramic video are relatively narrow, most of the hole wall rocks are basically the same, most of the hole wall rocks are basically free of special corner points, the image noise is very serious, and interference image signals formed by water flow sediment of the hole wall are relatively strong, the method and the device aim at the characteristics of the borehole hole wall images, and therefore a difference image rapid detection matching method based on ORB characteristics and BRIEF characteristic description operator principles is adopted to achieve characteristic detection and matching in the narrow-band images. The method for rapidly detecting and matching the difference image provided by the invention mainly comprises the following steps:
step 2.1, sequentially taking three adjacent frame narrow strip-shaped images, which are respectively defined as a first narrow strip-shaped image Img1, a second narrow strip-shaped image Img2 and a third narrow strip-shaped image Img 3; the first difference image ImgA is a difference value between the second narrowband image Img2 and the first narrowband image Img1, that is, ImgA is Img2 to Img 1; the second difference image ImgB is a difference value between the third narrowband image Img3 and the second narrowband image Img2, that is, ImgB is Img3 to Img 2. The difference image can well eliminate the inherent error of the image.
And 2.2, converting the first difference image ImgA into a first gray level image GrayA, and converting the second difference image ImgB into a second gray level image GrayB. Matching using grayscale images can provide speed and accuracy of operation.
And 2.3, detecting a first gray-scale image feature point and a first gray-scale image feature description operator (BRIEF feature description operator) in the first gray-scale image GrayA by utilizing an ORB algorithm, and detecting a second gray-scale image feature point and a second gray-scale image feature description operator (BRIEF feature description operator) in the second gray-scale image GrayB by utilizing the ORB algorithm. A plurality of feature points can be detected quickly in both the first gray image gray a and the second gray image gray b.
The characteristic points obtained by the ORB characteristic and BRIEF characteristic description operator adopted by the invention can be realized by using SURF, SIFT, FAST and Harris characteristic detection and characteristic matching methods under the condition of high image quality of the drilling video or low requirement on the drawing speed. Similar methods can refer to methods such as feature detection and feature matching of related images, for example, a method for feature detection and feature matching in OpenCV-blog of Jack _ Sarah-CSDN blog https:// block. Because the ORB method is relatively fast in operation speed and relatively stable, the invention preferably selects the ORB method to obtain the characteristic points of each frame of narrow-band borehole image in the invention. Under the conditions that video images are clear and real-time effect is not required, more stable and accurate feature point detection and matching can be realized by using feature detection and feature matching methods such as SURF and SIFT.
And 2.4, calculating the Hamming distance from each first gray image feature point to each second gray image feature point according to the first gray image feature description operator and the second gray image feature description operator by using the Hamming distance (Hamming).
Step 2.5, the minimum Hamming distance is an optimal matching distance, and a first gray image feature point and a second gray image feature point corresponding to the optimal matching distance are called as optimal feature matching point pairs (respectively marked as Ma and Mb); the sub-minimum Hamming distance is used as a sub-optimal matching distance, and the first gray image feature point and the second gray image feature point corresponding to the sub-optimal matching distance are called as sub-optimal feature matching point pairs (Ma 'and Mb');
the remaining pairs of characteristic points may be considered erroneous.
2.6, under the condition that the quality of the drilling image is good, directly adopting optimal feature matching point pairs (Ma and Mb) to express the image matching results of the first gray level image GrayA and the second gray level image GrayB; under the condition that the quality of the drilled hole image is poor and the effect of the formed narrow-band-shaped image is not ideal, the average value of the optimal feature matching point pair (Ma and Mb) and the suboptimal feature matching point pair (Ma 'and Mb') is used as an average feature matching point pair, and the image matching results of the first gray scale image GrayA and the second gray scale image GrayB are expressed. The final matching effect is shown in fig. 4.
Aiming at the particularity (narrow-band images are narrow, the longitudinal matching scale range is small, the transverse matching range is large, the local range has strong continuity and the image characteristics are seriously homogenized) and the specificity (only aiming at the narrow-band images obtained in the step (1)), because the noise interference of the borehole wall imaging is large, the rock wall structure in a certain range is similar and identical, and a large amount of different scale rotation and brightness change exist, the result of performing a large amount of matching on two adjacent frames of narrow-band images shows that: the method is a convenient and rapid optimal method for describing the up-down and left-right movement conditions of two adjacent narrow-band images by adopting the minimum feature matching distance.
And determining corresponding matched pixel points on the second narrow banded image Img2 and the third narrow banded image Img3 according to the first gray image characteristic point and the second gray image characteristic point corresponding to the optimal characteristic matching point pair or the average characteristic matching point pair.
Since the first gray scale image feature point corresponds to the first gray scale image gray a, and the first gray scale image gray a corresponds to the first narrow strip-shaped image Img1 and the second narrow strip-shaped image Img2, the corresponding matching pixel point on the second narrow strip-shaped image Img2 can be obtained through the first gray scale image feature point, and similarly, the corresponding matching pixel point on the third narrow strip-shaped image Img3 can be determined through the second gray scale image feature point.
The width matching offset in the width direction and the height matching offset in the height direction of the matching pixel points of the second narrowband image Img2 with respect to the matching pixel points on the third narrowband image Img3 are calculated.
Through the steps, the width matching offset and the height matching offset of the narrow-band image corresponding to the drilling image of each frame in the width direction and the height matching offset in the height direction relative to the matching pixel point of the narrow-band image of the drilling image of the next frame can be obtained, and the obtained narrow-band image is recorded in the characteristic file mData _ N.txt, except for the drilling image of the first frame which needs to be discarded.
And 3, calculating Azimuth pixel measurement offset and Depth pixel measurement offset according to the Azimuth data Azimuth and the Depth data Depth of the Depth encoder, and determining width actual offset and height actual offset.
And 3.1, acquiring Azimuth data Azimuth and Depth data Depth of each frame of narrow-band image corresponding to the second frame of borehole image in the in-hole panoramic video according to the acquired information file acquired in the step 1, firstly, forming the Azimuth data Azimuth and the Depth data Depth into a data array, performing smooth filtering processing on the data array by adopting a smooth filtering method, eliminating abnormal values in the Azimuth data Azimuth and the Depth data Depth, and enabling the finally acquired adjacent data in the Azimuth data Azimuth and the Depth data Depth to be in relatively smooth transition. The step aims to eliminate abnormal values in the collected information file and enable each group of data to be in relatively smooth fluctuation transition.
Then, calculating Azimuth value Azimuth of narrow-band image of current frameiAzimuth value Azimuth of narrow banded image with next framei+1Is used as the Azimuth measurement offset delta Azimuth of the narrow-band image of the current frameiI is the number of frames of the narrow band image,
calculating Depth data Depth of current frame narrow-band imageiDepth data Depth of narrow banded image with next framei+1Is used as the Depth measurement offset delta Depth of the narrow-band image of the current frameiI is the number of frames of the narrow band image,
because the width of the narrow band-shaped image obtained from the same drilling video image is a fixed value W, the width of the narrow band-shaped image is equal to the width of the narrow band-shaped imageIs the width of the narrow banded image, the azimuthal measurement offset Δ Azimuth of the narrow banded image of the current frameiCorresponding azimuthal pixel measurement offset is Δ wi＝ΔAzimuthi*W/360＝(Azimuthi+1-Azimuthi)*W/360。
In addition, according to the precision of the encoder and the general paying-off walking speed of an exploration hole of 72 m/h, the Depth measurement deviation delta Depth of the narrow strip image of the current frame can be obtainediCorresponding depth pixel measurement offset is Δ hi＝ΔDepthi*k。
k is the speed of paying off pixel points of high-definition drilling shooting, and the value of the k can be generally 2000 pixel points per meter.
Step 3.2, if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is smaller than W/4, the actual width offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is the width matching offset; if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is larger than or equal to W/4 and smaller than W, the actual width offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is an azimuth pixel measurement offset; if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is larger than or equal to W, the width actual offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is W.
If the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is smaller than a first height pixel threshold (the first height pixel threshold is smaller than H, preferably 0.5H), the actual height offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is the height matching offset; if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is greater than or equal to the first height pixel threshold and smaller than the second height pixel threshold (preferably, the second height pixel threshold is H), the actual height offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is a depth pixel measurement offset; and if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is greater than or equal to a second height pixel threshold value (H), the height actual offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is H.
Therefore, the effect of optimizing the correction of the matching data and the recorded data by mutual comparison is achieved, and the correctness of the matching data and the accuracy of the final image data are further improved.
Step 4, quickly splicing and fusing narrow banded images
Splicing the current frame narrow strip-shaped image and the next frame narrow strip-shaped image according to the width actual offset and the height actual offset, which specifically comprises the following steps:
circularly translating the current frame narrow-band image according to the actual width offset, correcting the offset of the current frame narrow-band image and the next frame narrow-band image in the width direction, performing weighted fusion on the overlapped part of the circularly translated current frame narrow-band image and the next frame narrow-band image in the height direction, and taking the pixel height of the overlapped part of the circularly translated current frame narrow-band image and the next frame narrow-band image in the height direction as the actual height offset. The weighted fusion refers to the sum of the first weighting coefficient multiplied by the pixel point on the current frame narrow-band image after circular translation and the second weighting coefficient multiplied by the pixel point corresponding to the next frame narrow-band image.
And in the direction from the current frame narrow-band-shaped image to the next frame narrow-band-shaped image, the first weighting coefficients of the overlapped parts of the current frame narrow-band-shaped image and the next frame narrow-band-shaped image in the height direction after circular translation are sequentially reduced, the second weighting coefficients are sequentially increased, and the sum of the first weighting coefficient and the second weighting coefficient is 1.
And (4) obtaining the splicing of the narrow banded images (except the narrow banded images of the frames corresponding to the first frame of the borehole images in the intra-hole panoramic video) of each frame according to the mode to obtain a fused spliced image.
The comparison graph of the effect of the image without weighted fusion (left image) and the fused and spliced image (right image) after the partial area is enlarged is shown in fig. 5.
In the process, due to the fact that the drill hole is deep, the video image data is very long and large, and in order to prevent the generated image from being too long to be stored for viewing, the fused and spliced image is divided into a plurality of fused and spliced sub-images according to the set depth length to be stored, in the embodiment, the actual drill hole investigation depth corresponding to 4360 narrow-band images of the example is 56.2 meters to 59.3 meters, the set depth length is 2 meters, and finally 2 fused and spliced sub-images are formed, as shown in fig. 6.
Step 5, optimizing the spliced image
And (4) aiming at the multiple fusion splicing subimages formed in the step (4), performing image enhancement on each fusion splicing subimage, wherein the image enhancement processing comprises gray scale stretching and gray scale equalization, and the processing effect of one fusion splicing subimage is shown in fig. 7. The purpose of image gray stretching is to prevent the image from being too dark or too bright, and uniform gray equalization processing (image histogram equalization image enhancement) is performed on each fusion splicing sub-image so that pixel transition is smooth.
In addition, in order to prevent the fused and spliced subimages from being too large or too long, the length and width pixel size conversion is carried out on the fused and spliced subimages subjected to the image enhancement processing. The image is more convenient to watch and understand and accords with aesthetic perception. The present invention employs an aspect ratio dimension of 0.618, close to the golden scale, for scaling the aspect ratio pixels.
Step 6, information labeling and histogram generation of the spliced image
And marking the Azimuth data Azimuth and the Depth data Depth of each fusion splicing subimage, namely, marking the Depth scale and identifying the south, east, west and north (N-E-S-W-N) Azimuth. Therefore, in order to facilitate engineering application, the drilling core diagram, namely the drilling histogram, is generated in the four directions of east, south, west and north according to the azimuth information of south, east and west and north. Taking a high definition video image as an example, a final effect map at a certain position after information annotation and histogram generation is performed is shown in fig. 8.
So far, the detailed implementation steps and the detailed process of the method are described, the method realizes the process of quickly matching and fusing the panoramic video images in the hole into the image, and provides great convenience for the actual drilling exploration process. The method can also realize express splicing and fusion of a large amount of high-definition video image data into a picture, and greatly improves the real-time performance and high efficiency of the shooting investigation in the deep rock mass structure hole.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.
Claims (7)
1. A method for quickly splicing and fusing annular images of panoramic videos in holes into an image is characterized by comprising the following steps:
step 1, obtaining each frame of drilling image obtained through an in-hole panoramic video to obtain each corresponding frame of narrow-band-shaped image, and obtaining Azimuth angle data Azimuth of a compass or an electronic compass in each frame of drilling image and Depth data Depth of a Depth encoder;
step 2, feature detection and matching, which specifically comprises the following steps:
step 2.1, sequentially taking three adjacent frame narrow strip-shaped images, which are respectively defined as a first narrow strip-shaped image Img1, a second narrow strip-shaped image Img2 and a third narrow strip-shaped image Img 3; the first difference image ImgA is a difference value between the second narrowband image Img2 and the first narrowband image Img1, that is, ImgA is Img2 to Img 1; the second difference image ImgB is a difference value of the third narrowband image Img3 and the second narrowband image Img2, that is, ImgB is Img3 to Img 2;
step 2.2, the first difference image ImgA is converted into a first gray level image gray a, the second difference image ImgB is converted into a second gray level image gray b,
step 2.3, detecting a first gray image feature point and a first gray image feature description operator in the first gray image GrayA, detecting a second gray image feature point and a second gray image feature description operator in the second gray image GrayB,
step 2.4, calculating the Hamming distance from each first gray image feature point to each second gray image feature point according to the first gray image feature description operator and the second gray image feature description operator,
step 2.5, the minimum Hamming distance is an optimal matching distance, and a first gray image characteristic point and a second gray image characteristic point corresponding to the optimal matching distance are called as optimal characteristic matching point pairs and are respectively marked as Ma and Mb; the sub-minimum Hamming distance is used as a sub-optimal matching distance, and a first gray image feature point and a second gray image feature point corresponding to the sub-optimal matching distance are called as sub-optimal feature matching point pairs and are respectively marked as Ma 'and Mb';
step 2.6, taking the average value of the optimal characteristic matching point pair and the suboptimal characteristic matching point pair as an average characteristic matching point pair,
determining corresponding matched pixel points on the second narrow banded image Img2 and the third narrow banded image Img3 according to the first gray image characteristic point and the second gray image characteristic point corresponding to the optimal characteristic matching point pair or the average characteristic matching point pair,
calculating a width matching offset in the width direction of the matching pixel points of the second narrowband image Img2 with respect to the matching pixel points on the third narrowband image Img3, and a height matching offset in the height direction,
step 3, calculating Azimuth pixel measurement offset and Depth pixel measurement offset of the narrowband image according to the Azimuth data Azimuth and the Depth data Depth of the Depth encoder, determining width actual offset and height actual offset of the narrowband image,
step 4, fusion splicing is carried out on the narrow strip image of the current frame and the narrow strip image of the next frame according to the actual width deviation and the actual height deviation,
the step 1 comprises the following steps:
identifying a center O (x, y) of the borehole image and a maximum radius Rmax and a minimum radius Rmin of a borehole wall annular image within the borehole image, wherein x and y are respectively an abscissa and an ordinate of the borehole image,
on the annular image of the hole wall of the drill hole, W pixel points are respectively collected on each circle with the radius of Rmin-Rmax, the W pixel points collected on each circle form corresponding pixel rows, a narrow-band-shaped image is generated according to each pixel row, the corresponding radius of the uppermost row to the lowermost row of the narrow-band-shaped image is from large to small, the width of the narrow-band-shaped image is W, the height of the narrow-band-shaped image is H, and H is Rmax-Rmin.
2. The method for rapidly splicing and fusing annular images of the intra-hole panoramic video according to claim 1, wherein the step 3 comprises the following steps:
step 3.1, obtaining Azimuth angle data Azimuth and Depth data Depth of each frame of narrow-band-shaped image, and calculating the Azimuth angle value Azimuth of the current frame of narrow-band-shaped imageiAzimuth value Azimuth of narrow banded image with next framei+1Is used as the Azimuth measurement offset delta Azimuth of the narrow-band image of the current frameiI is the number of frames of the narrow band image,
calculating Depth data Depth of current frame narrow-band imageiDepth data Depth of narrow banded image with next framei+1Is used as the Depth measurement offset delta Depth of the narrow-band image of the current framei，
Measuring the offset Δ Azimuth by the Azimuth of the narrow banded image of the current frameiAnd the Depth measurement offset Δ Depth of the current frame narrow-band imageiCalculate an azimuth pixel measurement offset and a depth pixel measurement offset,
step 3.2, if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is smaller than W/4, the actual width offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is the width matching offset; if the absolute value of the width matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is larger than or equal to W/4 and smaller than W, the actual width offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the width direction is an azimuth pixel measurement offset; if the absolute value of the width matching offset of the current frame narrow-band image relative to the next frame narrow-band image in the width direction is greater than or equal to W, the width actual offset of the current frame narrow-band image relative to the next frame narrow-band image in the width direction is W,
if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is smaller than the first height pixel threshold value, the actual height offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is the height matching offset; if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is greater than or equal to a first height pixel threshold value and smaller than a second height pixel threshold value, the actual height offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is a depth pixel measurement offset; and if the absolute value of the height matching offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is greater than or equal to the second height pixel threshold value, the height actual offset of the current frame narrow-band-shaped image relative to the next frame narrow-band-shaped image in the height direction is H.
3. The method for rapidly splicing and fusing annular images of an intra-hole panoramic video according to claim 2, wherein Azimuth data Azimuth and Depth data Depth of each frame of the narrow annular image in step 3.1 are both smoothed.
4. The method for rapidly splicing and fusing annular images of the intra-hole panoramic video according to claim 1, wherein the step 4 comprises the following steps:
circularly translating the current frame narrow-band image according to the actual width offset, correcting the offset of the current frame narrow-band image and the next frame narrow-band image in the width direction, performing weighted fusion on the overlapped part of the circularly translated current frame narrow-band image and the next frame narrow-band image in the height direction, taking the pixel height of the overlapped part of the circularly translated current frame narrow-band image and the next frame narrow-band image in the height direction as the actual height offset, obtaining a spliced fusion splicing image of each frame narrow-band image, and dividing the fusion splicing image into a plurality of fusion splicing sub-images according to the set depth length.
5. The method as claimed in claim 4, wherein in a direction from a current frame of narrow-band image to a next frame of narrow-band image, a first weighting coefficient of a portion of the circularly translated current frame of narrow-band image overlapped with the next frame of narrow-band image in a height direction is sequentially decreased, a second weighting coefficient is sequentially increased, and the sum of the first weighting coefficient and the second weighting coefficient is 1.
6. The method for rapidly splicing and fusing annular images of the intra-hole panoramic video according to claim 4, further comprising the step 5: and 4, performing image enhancement and length-width pixel size conversion on each of the multiple fusion splicing subimages formed in the step 4.
7. The method for rapidly splicing and fusing annular images of the intra-hole panoramic video according to claim 6, further comprising the step 6: and marking Azimuth data Azimuth and Depth data Depth of each fusion splicing subimage, and generating a drilling histogram in four directions of east, south, west and north respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910903069.8A CN110689512B (en) | 2019-09-24 | 2019-09-24 | Method for quickly splicing and fusing annular images of panoramic video in hole into image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910903069.8A CN110689512B (en) | 2019-09-24 | 2019-09-24 | Method for quickly splicing and fusing annular images of panoramic video in hole into image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110689512A CN110689512A (en) | 2020-01-14 |
| CN110689512B true CN110689512B (en) | 2022-03-08 |
Family
ID=69110040
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910903069.8A Active CN110689512B (en) | 2019-09-24 | 2019-09-24 | Method for quickly splicing and fusing annular images of panoramic video in hole into image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110689512B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113077488B (en) * | 2021-04-02 | 2022-07-01 | 昆明理工大学 | River surface flow velocity detection method and device |
| CN114972019B (en) * | 2021-04-14 | 2024-05-24 | 华东师范大学 | Depth image stitching method and device based on TOF camera and computer equipment |
| CN114584747B (en) * | 2022-03-04 | 2023-10-31 | 大连海事大学 | 360-degree annular curtain seamless projection soft correction method |
| CN114511048B (en) * | 2022-04-20 | 2022-07-29 | 深圳市海清视讯科技有限公司 | Image processing method, device, equipment and system |
| CN115861588A (en) * | 2022-12-26 | 2023-03-28 | 贵州大学 | Coal mine fault identification method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6750860B1 (en) * | 1998-12-28 | 2004-06-15 | Microsoft Corporation | Rendering with concentric mosaics |
| CN103517041A (en) * | 2013-09-29 | 2014-01-15 | 北京理工大学 | Real-time full-view monitoring method and device based on multi-camera rotating scanning |
| CN105957015A (en) * | 2016-06-15 | 2016-09-21 | 武汉理工大学 | Thread bucket interior wall image 360 DEG panorama mosaicing method and system |
| CN107424120A (en) * | 2017-04-12 | 2017-12-01 | 湖南源信光电科技股份有限公司 | A kind of image split-joint method in panoramic looking-around system |
-
2019
- 2019-09-24 CN CN201910903069.8A patent/CN110689512B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6750860B1 (en) * | 1998-12-28 | 2004-06-15 | Microsoft Corporation | Rendering with concentric mosaics |
| CN103517041A (en) * | 2013-09-29 | 2014-01-15 | 北京理工大学 | Real-time full-view monitoring method and device based on multi-camera rotating scanning |
| CN105957015A (en) * | 2016-06-15 | 2016-09-21 | 武汉理工大学 | Thread bucket interior wall image 360 DEG panorama mosaicing method and system |
| CN107424120A (en) * | 2017-04-12 | 2017-12-01 | 湖南源信光电科技股份有限公司 | A kind of image split-joint method in panoramic looking-around system |
Non-Patent Citations (2)
| Title |
|---|
| A system for measuring borehole diametric deformation based on mechanical contact and micro-optical imaging;Chuanying Wang et al.;《Measurement》;20180731;191-197 * |
| 基于全景相机组的立体视觉系统设计与实现;郑轩;《中国优秀硕士学位论文全文数据库 (信息科技辑)》;20180415;I138-2653 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110689512A (en) | 2020-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110689512B (en) | Method for quickly splicing and fusing annular images of panoramic video in hole into image | |
| CN106485751B (en) | Unmanned aerial vehicle photographic imaging and data processing method and system applied to foundation pile detection | |
| CN103440629B (en) | Laser labelling is from the digital image processing method of the Video Extensometer of motion tracking | |
| CN109409290B (en) | Thermometer verification reading automatic identification system and method | |
| CN106657789A (en) | Thread panoramic image synthesis method | |
| CN107807387A (en) | Acquisition methods when seismic first break based on neutral net is walked | |
| CN113610060B (en) | Structure crack sub-pixel detection method | |
| CN113689401A (en) | Method and device for detecting diameter of crystal bar of czochralski silicon single crystal furnace | |
| CN117570881B (en) | Land area measurement system and method for investigation of homeland resources | |
| US20160307062A1 (en) | Method and device for determining regions to be detected on display motherboard | |
| CN109376641A (en) | A kind of moving vehicle detection method based on unmanned plane video | |
| CN106056121A (en) | Satellite assembly workpiece fast-identification method based on SIFT image feature matching | |
| CN113569647A (en) | AIS-based ship high-precision coordinate mapping method | |
| Zou et al. | The fast formation of high-precision panoramic image for the processing of borehole camera video of deep rock mass structures | |
| CN106296678A (en) | RQD based on boring optical image technology analyzes method | |
| CN113807238B (en) | Visual measurement method for river channel water surface floater area | |
| CN104156977A (en) | Point target movement velocity detection method based on multiple linear moveout scanning, extending and sampling | |
| CN112161581B (en) | Machine vision shield tail clearance dynamic measurement method considering segment end surface deflection angle | |
| CN113689451A (en) | Carbon plate boundary extraction method and device, storage medium and electronic equipment | |
| CN110728691B (en) | Multi-temporal water sideline-based coastline automatic judgment method | |
| CN112116644A (en) | Vision-based obstacle detection method and device and obstacle distance calculation method and device | |
| CN110533698B (en) | Foundation pit construction pile detection control method based on visual detection | |
| CN112164118A (en) | Geographic image processing system and method | |
| CN111724340A (en) | Farmland margin line visual detection method and system | |
| CN110443844B (en) | Rock mass structural plane geometric information extraction method based on drilling image |
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 |