CN114677340A - Concrete surface roughness detection method based on image edge - Google Patents

Concrete surface roughness detection method based on image edge Download PDF

Info

Publication number
CN114677340A
CN114677340A CN202210246723.4A CN202210246723A CN114677340A CN 114677340 A CN114677340 A CN 114677340A CN 202210246723 A CN202210246723 A CN 202210246723A CN 114677340 A CN114677340 A CN 114677340A
Authority
CN
China
Prior art keywords
image
edge
pixel
roughness
concrete
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.)
Granted
Application number
CN202210246723.4A
Other languages
Chinese (zh)
Other versions
CN114677340B (en
Inventor
左健存
马佳军
李光洁
詹强
吴丹丹
常远培
薛颖
张宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Polytechnic University
Original Assignee
Shanghai Polytechnic University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Polytechnic University filed Critical Shanghai Polytechnic University
Priority to CN202210246723.4A priority Critical patent/CN114677340B/en
Priority claimed from CN202210246723.4A external-priority patent/CN114677340B/en
Publication of CN114677340A publication Critical patent/CN114677340A/en
Application granted granted Critical
Publication of CN114677340B publication Critical patent/CN114677340B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/73—Deblurring; Sharpening
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/181—Segmentation; Edge detection involving edge growing; involving edge linking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • 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
    • G06T2207/20024—Filtering details
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30132—Masonry; Concrete

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a method for detecting the roughness of a concrete surface based on an image edge; the detection method comprises the following steps: 1) acquiring images of the surface of a sample block with a known concrete roughness value; 2) preprocessing the acquired image; 3) graying the preprocessed color image; 4) processing the gray level image; 5) carrying out pixel level and sub-pixel level edge detection on the processed gray level image; 6) performing logical OR operation on the edge image; 7) calculating edge frequency; 8) fitting an edge frequency and roughness curve by a least square method; and (3) repeating the steps 1) to 7) for the concrete image with unknown roughness, and calculating the edge frequency to obtain the roughness value of the concrete image. The method can be used for detecting the surface roughness of the concrete under portable mobile phone equipment, is simple and efficient, and achieves higher precision.

Description

Concrete surface roughness detection method based on image edge
Technical Field
The invention belongs to the technical field of image processing, and particularly relates to a concrete surface roughness detection method based on image edges.
Background
The pouring of concrete of different ages to newly build, repair or reform concrete structures requires sufficient adhesion between the concrete castings. The bond strength between different castings is attributed to the surface roughness. The surface roughness may be achieved in a variety of ways, such as sandblasting, napping, printing, and the like. At present, common detection methods for concrete surface roughness include a sand-pouring method, a mechanical probe method, a profilometer method and the like. In the face of thousands of concrete blocks to be tested, the traditional methods for detecting the rough surface of the concrete are time-consuming, complex to operate, not easy to carry operating equipment and expensive in detection cost.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a method for detecting the surface roughness of the concrete based on the image edge, which can simply and efficiently detect the concrete under a portable mobile phone device and reach higher precision.
The invention adopts the following technical scheme:
a method for detecting the roughness of a concrete surface based on an image edge comprises the following steps:
step 1: image acquisition
Acquiring an image of the surface of a sample block with a known concrete roughness value, wherein the roughness value is detected by a three-dimensional laser scanner;
And 2, step: color image pre-processing
Preprocessing the color image, and correcting the problems of uneven brightness and unclear image caused by uneven illumination of the image;
and step 3: graying of color images
For the preprocessed color image, calculating the weighted sum of each pixel R, G and B component, and converting the pixel RGB value into a gray value to obtain a gray image;
and 4, step 4: gray scale image processing
Carrying out mean filtering on the gray level image to obtain a filtered image, and realizing the effect of smoothing noise by using the mean value of pixels around a certain pixel point; secondly, performing saturation processing on the lowest 1% and the highest 1% of all pixel values so as to improve the contrast of an output image;
and 5: pixel-level and sub-pixel-level edge detection
Respectively adopting Canny operators to realize pixel level edge detection on the gray level image processed in the step 4, and realizing sub-pixel level edge detection by using Zernike moments to respectively obtain pixel level edge images and sub-pixel level edge images;
step 6: edge image logical OR operation
Carrying out logic OR operation on the edge images detected by the Canny edge detection and the subpixel level Zernike edge detection;
and 7: calculating edge frequencies
Defining the ratio of the white edge pixel point number to the total pixel in the new image after logical OR operation as an edge frequency, and assuming that the image resolution is X Y, the edge frequency EF is:
Figure BDA0003545345800000021
wherein N (1) is a white pixel, and N (0) is a black pixel;
and step 8: fitting edge frequency and roughness curves using least squares
Fitting an edge frequency curve and a roughness curve by adopting a least square method based on all the acquired roughness data of the concrete image with known surface roughness and corresponding edge frequency data obtained by calculation;
further, for the concrete image with unknown roughness, only the steps 2 to 7 need to be repeated, the edge frequency is calculated, and then the corresponding roughness value is obtained based on the curve.
In the invention, in step 2, the method for preprocessing the color image is as follows:
firstly, extracting an illumination variable through bright-pass bilateral filtering, converting an RGB three-channel color image into an HSV channel, and estimating an illumination component from a V channel;
and then constructing a two-dimensional Gama gamma function, adjusting parameters of the two-dimensional gamma function by utilizing the illumination distribution characteristics, reducing the brightness value of the over-illumination strong area, improving the brightness value of the image in the over-illumination dark area, and realizing the self-adaptive correction of the image with uneven illumination.
Compared with the prior art, the invention has the beneficial effects that:
the invention can adaptively correct the image with uneven illumination by an image processing method under the portable mobile phone equipment, obtains more complete image edges by utilizing two-stage edge detection and performing logic OR operation, and provides a new edge frequency corresponding to the roughness to perform simple and efficient detection. More unknown concrete surface roughness is predicted through the concrete surface image with limited known roughness, and the detection accuracy is higher.
Drawings
FIG. 1 is an image of the rough surface of the original concrete of the present invention.
Fig. 2 is a pre-processed image of an original color image according to the present invention.
Fig. 3 is an image after the edge detection using Canny according to the present invention.
Figure 4 is an image of the invention after detection of edges using Zernike moments.
Figure 5 is an image of the present invention after logical or' ing Canny and Zernike moments.
FIG. 6 is a graph of edge frequency versus roughness that was fitted in accordance with the present invention.
Fig. 7 is a graph of the deviation of the detected image value from the actual value according to the present invention.
Fig. 8 is a graph of deviation of a detected value from an actual value based on Canny edges.
Figure 9 is a graph of deviation of detected and actual values based on Zernike sub-pixel edges.
Detailed Description
The invention is further illustrated by the following figures and examples, which should not be construed as limiting the scope of the invention.
The invention provides a method for detecting the roughness of a concrete surface based on an image edge, which comprises the following steps:
step 1: image acquisition
The method comprises the steps of utilizing an android smart phone with more than 1200 ten thousand pixels to carry out image acquisition on the surface of a sample block with a known concrete roughness value, wherein the roughness value is detected by a three-dimensional laser scanner.
And 2, step: color image pre-processing
Because the light source is unevenly irradiated on the rough surface, an image with uniform illumination is obtained. Firstly, extracting illumination variables through bright-pass bilateral filtering, converting an RGB three-channel color image into HSV channels, and estimating illumination components g (i) from a V channel:
Figure BDA0003545345800000031
Figure BDA0003545345800000032
Figure BDA0003545345800000033
Figure BDA0003545345800000034
where f (i) denotes the V channel at the i pixel location, the spatial kernel ω (i) is Gaussian,
Figure BDA0003545345800000035
is single-sided Gaussian, theta and sigma are the respective standard deviation, and the omega range is [ -omega, + omega]2。
A two-dimensional Gama gamma function is constructed, parameters of the two-dimensional gamma function are adjusted by utilizing the illumination distribution characteristics, the brightness value of an over-illumination area is reduced, the brightness value of an image in an over-illumination dark area is improved, and self-adaptive correction of an image with uneven illumination is realized;
Figure BDA0003545345800000036
Figure BDA0003545345800000037
Wherein F (i) is an input RGB image, g (i) is an illumination component,
Figure BDA0003545345800000038
the two-dimensional gamma function o (i) achieves non-uniformity correction of illumination by adjusting the enhancement index γ for the luminance mean of the illumination component.
And 3, step 3: graying of color images
For the processed color image, the pixel RGB values are converted to gray scale values by calculating the weighted sum of each pixel R, G and the B component, resulting in a gray scale image:
gray＝0.2989*R+0.5870*G+0.1140*B
and 4, step 4: gray scale image processing
Carrying out 3-to-3 mean filtering on the gray level image to obtain a filtered image u (x, y), and realizing the effect of smoothing noise by using the mean value of pixels around a certain pixel point; secondly, performing saturation processing on the lowest 1% and the highest 1% of all pixel values, setting the gray value smaller than 255 x 0.01 as 0, and setting the gray value larger than 255 x 0.99 as 255, so as to improve the contrast of the output image;
Figure BDA0003545345800000041
wherein p (S, t) represents a grayscale image, SxyA filtering window with the size of 3 x 3 and a central point at (x, y), and u (x, y) represents an image obtained after mean filtering;
and 5: pixel-level and sub-pixel-level edge detection
Pixel-level Canny operator edge detection:
smooth image with Gaussian filter
Convolving the two-dimensional Gaussian kernel with the image after gray processing, wherein a two-dimensional Gaussian function G (x, y):
Figure BDA0003545345800000042
(ii) Sobel operator calculates pixel gradient
Figure BDA0003545345800000043
Figure BDA0003545345800000044
Figure BDA0003545345800000045
Figure BDA0003545345800000046
Wherein S isxAnd SySobel operator 3 x 3, I is a gray image matrix, GxAnd GyGradient matrix in x and y directions, M (x, y) gradient magnitude matrix, and α (x, y) gradient magnitude matrixAn angle image.
Non-maximum pixel gradient suppression
Taking each point in the gradient amplitude matrix M (x, y) as a central pixel point, firstly finding out the direction d closest to alpha (x, y)k. If M (x, y) is along dkIf the direction is larger than the two adjacent pixel values, let gN(x, y) M (x, y), otherwise gN(x, y) is 0, in which case gN(x, y) is the non-maximum inhibition map.
Double-threshold detection for determining real edge position edge
Setting the high threshold to THThe low threshold is TLThe dual-threshold processing is regarded as the superposition of two threshold images:
gNH(x,y)＝gN(x,y)≥TH
gNL(x,y)＝gN(x,y)≥TL
wherein, gNH(x, y) is an image with amplitude greater than a high threshold value in the non-maximum image; gNL(x, y) is the image with amplitude greater than the low threshold value in the non-maximum image. And g isNL(x, y) includes gNHAll non-zero pixels in (x, y). By the formula, from gNLAll the landing g in (x, y) is deletedNHNon-zero pixel points of (x, y).
gNL(x,y)＝gNL(x,y)-gNH(x,y)
Subpixel level Zernike moment (Zernike) edge detection:
the sub-pixel edge location principle based on the Zernike moments is that 4 parameters of a model, background gray level h, step gray level k, vertical distance l from the center to the edge, and angle between the vertical line of the edge and the x axis are calculated through three defined Zernike moments A00, A11 and A20 with different orders
Figure BDA0003545345800000051
Then comparing the parameters with a set threshold value so as to accurately position the edge of the image; the m-th order n-zernike moments of f (x, y) for successive images are defined as:
Figure BDA0003545345800000052
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003545345800000053
is an integral kernel function;
after rotating the image, the image is symmetric about the x-axis:
Figure BDA0003545345800000054
Figure BDA0003545345800000055
where f' (x, y) is the rotated image.
The angle of rotation φ can be calculated by the formula:
Figure BDA0003545345800000056
rotational invariance is an important property of the zernike moment, which is the image after rotation phi:
Figure BDA0003545345800000057
Figure BDA0003545345800000058
Figure BDA0003545345800000061
then l, k, h can be calculated, and then the edge positions of the sub-pixels are:
Figure BDA0003545345800000062
where (x, y) is the pixel level location.
In consideration of the template effect, deviations will occur in the calculation of the actual sub-pixels, and assuming that the template is M × M, the sub-pixel coordinate formula is modified as follows:
Figure BDA0003545345800000063
step 6: edge image logical OR operation
C is the edge image detected by Canny, Z is the edge image detected by Zernike, and the logical OR operation is carried out on the C and the Z.
Q＝C|Z
Wherein Q is a new image generated after logical OR operation;
and 7: calculating edge frequencies
Defining the ratio of the white edge pixel points in the image to the total pixels of the image as an edge frequency, and assuming that the resolution of the image is X Y, the edge frequency EF is:
Figure BDA0003545345800000064
wherein N (1) is a white pixel and N (0) is a black pixel.
And step 8: least squares fitting of edge frequency and roughness curves
The least squares method finds the best match for the data by minimizing the sum of squared errors, for a given k sets of observations
Data (x)i,yi) (k ═ 1,2, … k), solving an objective function:
Figure BDA0003545345800000065
wherein x is [ x ]i,xi,…xi]TFor the calculated edge frequency EF, ω ═ ω [ ω ]1,ω2,…ωi]TFor the parameters to be determined, y ═ h (x, w) is the roughness value detected by the three-dimensional laser scanner.
After the edge frequency is calculated for all the collected concrete images with known surface roughness and curve fitting is completed, for the concrete images with unknown roughness, the steps 1 to 7 are repeated only, and the edge frequency is calculated, so that a roughness value in the curve corresponds to the edge frequency.
In a specific embodiment, the concrete surface roughness is detected by the detection method:
firstly, according to the step 1, image acquisition is carried out on the surface of a sample block with a known concrete roughness value by using an android smart phone with more than 1200 ten thousand pixels, and 9 pictures are acquired in total. One image is shown in fig. 1 and the roughness value is detected by the three-dimensional laser scanner as 2.1999.
Next, the color image is preprocessed in step 2, the illumination component is extracted by bright-pass bilateral filtering, and the image is adaptively corrected for illumination non-uniformity using a gamma function, as shown in fig. 2.
Next, the color image is grayed out and the grayscale image is processed in steps 3 and 4. The amount of information in a color image is too large, and the complexity of image processing is reduced by graying.
Next, pixel-level Canny edge detection and sub-pixel-level Zernike edge detection are taken for the gray-scale image, per step 5. Detected edge maps at the pixel level and sub-pixel level, as shown in fig. 3 and 4.
Then, according to step 6, performing logical or operation on the edge image obtained by Canny edge detection and subpixel level Zernike edge detection, as shown in fig. 5;
the edge frequency is calculated for the new image after the logical or operation, according to step 7.
Finally, according to step 8, the edge frequency and roughness curves are fitted by using the least square method, and the fitted curves are shown in fig. 6.
After the edge frequency is calculated for all the collected concrete images with known surface roughness and curve fitting is completed, for the concrete images with unknown roughness, the steps 1 to 7 are repeated only, and the edge frequency is calculated, so that a roughness value in the curve corresponds to the edge frequency. 6 unknown images are shot, corresponding roughness values are found by calculating edge frequency, and the detection precision is over 94.7% by comparing actual roughness values as shown in figure 7. If only Canny edge detection or sub-pixel edge detection of Zernike matrix is performed in step 6, many detailed edge information can not be extracted, resulting in measurement inaccuracy, and the result is shown in fig. 8 and 9. The results show that the accuracy of the measurement method proposed herein is improved by nearly 30% compared to the calculated edge frequency for a single Canny edge test and fitting a relationship curve to roughness. And the accuracy of the measurement method is improved by nearly 7% by a relation curve which is fit with the edge frequency calculated based on the Zernike matrix.
While the present invention has been described in detail with reference to the preferred embodiments, it should be understood that the above description should not be taken as limiting the invention. Various modifications and alterations to this invention will become apparent to those skilled in the art upon reading the foregoing description. Accordingly, the scope of the invention should be determined from the following claims.

Claims (3)

1. A detection method of concrete surface roughness based on image edges is characterized by comprising the following steps:
step 1: image acquisition
Acquiring an image of the surface of a sample block with a known concrete roughness value, wherein the roughness value is detected by a three-dimensional laser scanner;
step 2: color image preprocessing
Preprocessing the color image, and correcting the problems of uneven brightness and unclear image caused by uneven illumination of the image;
and step 3: graying of color images
For the preprocessed color image, calculating the weighted sum of each pixel R, G and B component, and converting the pixel RGB value into a gray value to obtain a gray image;
and 4, step 4: gray scale image processing
Carrying out mean filtering on the gray level image to obtain a filtered image, and realizing the effect of smoothing noise by using the mean value of pixels around a certain pixel point; secondly, performing saturation processing on the lowest 1% and the highest 1% of all pixel values so as to improve the contrast of an output image;
And 5: pixel-level and sub-pixel-level edge detection
Respectively adopting Canny operators to realize pixel level edge detection on the gray level image processed in the step 4, and realizing sub-pixel level edge detection by using Zernike moments to respectively obtain pixel level edge images and sub-pixel level edge images;
step 6: edge image logical OR operation
Carrying out logic OR operation on the edge images detected by the Canny edge detection and the Zernike sub-pixel edge detection;
and 7: calculating edge frequencies
In the new image after logical OR operation, the ratio of the white edge pixel point number to the total pixel is defined as the edge frequency,
assuming that the image resolution is X Y, the edge frequency EF is:
Figure FDA0003545345790000011
wherein N (1) is a white pixel and N (0) is a black pixel;
and 8: fitting edge frequency and roughness curves using least squares
Fitting a relation curve of the edge frequency and the roughness by adopting a least square method based on all the acquired roughness data of the concrete image with known surface roughness and the corresponding edge frequency data obtained by calculation; further, for the concrete image with unknown roughness, only the steps 2 to 7 need to be repeated, the edge frequency is calculated, and then the corresponding roughness value is obtained through the relation curve.
2. The detection method according to claim 1, wherein in step 2, the color image is preprocessed as follows:
firstly, extracting an illumination variable through bright-pass bilateral filtering, converting an RGB three-channel color image into an HSV channel, and estimating an illumination component from a V channel;
and then constructing a two-dimensional Gama gamma function, adjusting parameters of the two-dimensional gamma function by utilizing the illumination distribution characteristics, reducing the brightness value of the over-illumination strong area, improving the brightness value of the image in the over-illumination dark area, and realizing the self-adaptive correction of the image with uneven illumination.
3. The method according to claim 1, wherein in step 6, the Zernike moments are used to perform edge detection at sub-pixel level as follows:
it is to calculate 4 parameters of the model, background gray level h, step gray level k, vertical distance l from center to edge, angle between the vertical line of edge and x-axis by three defined Zernike moments A00, A11 and A20 with different orders
Figure FDA0003545345790000021
Then comparing the parameters with a set threshold value so as to accurately position the edge of the image;
the nth zernike moment of order m of f (x, y) of successive images is defined as:
Figure FDA0003545345790000022
wherein the content of the first and second substances,
Figure FDA0003545345790000023
is an integral kernel function;
after rotating the image, the image is symmetric about the x-axis:
Figure FDA0003545345790000024
Figure FDA0003545345790000025
Where f' (x, y) is the rotated image;
the angle of rotation phi is calculated by the formula:
Figure FDA0003545345790000026
rotational invariance is an important property of the zernike moment, which is the image after rotation phi:
Figure FDA0003545345790000027
Figure FDA0003545345790000028
Figure FDA0003545345790000029
then, l, k, h are calculated, and the edge positions of the sub-pixels are:
Figure FDA0003545345790000031
where (x, y) is the pixel level location.
Considering the template effect, there will be a deviation in the actual calculation of the sub-pixels, and assuming that the template is M × M, the sub-pixel coordinate formula is modified as:
Figure FDA0003545345790000032
CN202210246723.4A 2022-03-14 Concrete surface roughness detection method based on image edge Active CN114677340B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210246723.4A CN114677340B (en) 2022-03-14 Concrete surface roughness detection method based on image edge

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210246723.4A CN114677340B (en) 2022-03-14 Concrete surface roughness detection method based on image edge

Publications (2)

Publication Number Publication Date
CN114677340A true CN114677340A (en) 2022-06-28
CN114677340B CN114677340B (en) 2024-05-24

Family

ID=

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114882045A (en) * 2022-07-11 2022-08-09 山东金三星机械有限公司 Technological method for milling casting gate
CN117197534A (en) * 2023-08-04 2023-12-08 广州电缆厂有限公司 Automatic detection method for cable surface defects based on feature recognition

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20010020950A1 (en) * 2000-02-25 2001-09-13 International Business Machines Corporation Image conversion method, image processing apparatus, and image display apparatus
US20060257048A1 (en) * 2005-05-12 2006-11-16 Xiaofan Lin System and method for producing a page using frames of a video stream
WO2009072537A1 (en) * 2007-12-04 2009-06-11 Sony Corporation Image processing device and method, program, and recording medium
US8116522B1 (en) * 2008-08-25 2012-02-14 The United States Of America As Represented By The Secretary Of The Navy Ship detection system and method from overhead images
WO2015077493A1 (en) * 2013-11-20 2015-05-28 Digimarc Corporation Sensor-synchronized spectrally-structured-light imaging
CN104732536A (en) * 2015-03-18 2015-06-24 广东顺德西安交通大学研究院 Sub-pixel edge detection method based on improved morphology
CN106767564A (en) * 2016-11-03 2017-05-31 广东工业大学 A kind of detection method for being applied to phone housing surface roughness
WO2019041590A1 (en) * 2017-08-31 2019-03-07 中国科学院微电子研究所 Edge detection method using arbitrary angle
CN110870007A (en) * 2017-03-31 2020-03-06 弗劳恩霍夫应用研究促进协会 Apparatus and method for determining predetermined characteristics related to artificial bandwidth limiting processing of audio signals
WO2020082593A1 (en) * 2018-10-26 2020-04-30 深圳市华星光电技术有限公司 Method and device for enhancing image contrast
KR20200089410A (en) * 2019-01-17 2020-07-27 정인호 Low-light image correction method based on optimal gamma correction

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20010020950A1 (en) * 2000-02-25 2001-09-13 International Business Machines Corporation Image conversion method, image processing apparatus, and image display apparatus
US20060257048A1 (en) * 2005-05-12 2006-11-16 Xiaofan Lin System and method for producing a page using frames of a video stream
WO2009072537A1 (en) * 2007-12-04 2009-06-11 Sony Corporation Image processing device and method, program, and recording medium
US8116522B1 (en) * 2008-08-25 2012-02-14 The United States Of America As Represented By The Secretary Of The Navy Ship detection system and method from overhead images
WO2015077493A1 (en) * 2013-11-20 2015-05-28 Digimarc Corporation Sensor-synchronized spectrally-structured-light imaging
CN104732536A (en) * 2015-03-18 2015-06-24 广东顺德西安交通大学研究院 Sub-pixel edge detection method based on improved morphology
CN106767564A (en) * 2016-11-03 2017-05-31 广东工业大学 A kind of detection method for being applied to phone housing surface roughness
CN110870007A (en) * 2017-03-31 2020-03-06 弗劳恩霍夫应用研究促进协会 Apparatus and method for determining predetermined characteristics related to artificial bandwidth limiting processing of audio signals
WO2019041590A1 (en) * 2017-08-31 2019-03-07 中国科学院微电子研究所 Edge detection method using arbitrary angle
WO2020082593A1 (en) * 2018-10-26 2020-04-30 深圳市华星光电技术有限公司 Method and device for enhancing image contrast
KR20200089410A (en) * 2019-01-17 2020-07-27 정인호 Low-light image correction method based on optimal gamma correction

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
刘晓刚;莫毅;: "一种基于焊缝特征的焊缝中心线信息获取方法", 电焊机, no. 07, 20 July 2010 (2010-07-20) *
吴月波;徐晨;董蓉;: "基于灰度梯度和反正切拟合的亚像素边缘检测", 电视技术, no. 22, 17 November 2015 (2015-11-17) *
郎晓萍;刘力双;: "图像测量中快速边缘亚像素定位研究", 工具技术, no. 03, 20 March 2009 (2009-03-20) *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114882045A (en) * 2022-07-11 2022-08-09 山东金三星机械有限公司 Technological method for milling casting gate
CN114882045B (en) * 2022-07-11 2022-09-20 山东金三星机械有限公司 Technological method for milling casting gate
CN117197534A (en) * 2023-08-04 2023-12-08 广州电缆厂有限公司 Automatic detection method for cable surface defects based on feature recognition
CN117197534B (en) * 2023-08-04 2024-04-05 广州电缆厂有限公司 Automatic detection method for cable surface defects based on feature recognition

Similar Documents

Publication Publication Date Title
CN109872309B (en) Detection system, method, device and computer readable storage medium
CN108846397B (en) Automatic detection method for cable semi-conducting layer based on image processing
CN112651968B (en) Wood board deformation and pit detection method based on depth information
CN114897864B (en) Workpiece detection and defect judgment method based on digital-analog information
CN109409290B (en) Thermometer verification reading automatic identification system and method
CN112767359B (en) Method and system for detecting corner points of steel plate under complex background
CN104899888B (en) A kind of image sub-pixel edge detection method based on Legendre squares
CN115170669B (en) Identification and positioning method and system based on edge feature point set registration and storage medium
CN115096206B (en) High-precision part size measurement method based on machine vision
CN111539927B (en) Detection method of automobile plastic assembly fastening buckle missing detection device
CN107622475A (en) Gray correction method in image mosaic
CN113252103A (en) Method for calculating volume and mass of material pile based on MATLAB image recognition technology
CN111553927B (en) Checkerboard corner detection method, detection system, computer device and storage medium
CN113538399A (en) Method for obtaining accurate contour of workpiece, machine tool and storage medium
CN112801141A (en) Heterogeneous image matching method based on template matching and twin neural network optimization
CN114677340B (en) Concrete surface roughness detection method based on image edge
CN114677340A (en) Concrete surface roughness detection method based on image edge
JP2002090308A (en) Evaluation system for degree of surface degradation of steel using image processing
CN114998571B (en) Image processing and color detection method based on fixed-size markers
CN105005985B (en) Backlight image micron order edge detection method
CN114219802B (en) Skin connecting hole position detection method based on image processing
CN113763380B (en) Vector gradient-based reference-free image definition evaluation method
TW200842339A (en) Mura detection method and system
CN113469171A (en) Method, device and medium for identifying interest area in SFR test card image
CN116309760B (en) Cereal image alignment method and cereal detection equipment

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