CN111047556B - Strip steel surface defect detection method and device - Google Patents
Strip steel surface defect detection method and device Download PDFInfo
- Publication number
- CN111047556B CN111047556B CN201911107307.0A CN201911107307A CN111047556B CN 111047556 B CN111047556 B CN 111047556B CN 201911107307 A CN201911107307 A CN 201911107307A CN 111047556 B CN111047556 B CN 111047556B
- Authority
- CN
- China
- Prior art keywords
- image
- strip steel
- steel surface
- defect
- defects
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 184
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 131
- 239000010959 steel Substances 0.000 title claims abstract description 131
- 238000001514 detection method Methods 0.000 title claims abstract description 106
- 238000003672 processing method Methods 0.000 claims abstract description 11
- 238000000034 method Methods 0.000 claims description 34
- 238000012545 processing Methods 0.000 claims description 24
- 230000009466 transformation Effects 0.000 claims description 10
- 238000003708 edge detection Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 4
- 230000001629 suppression Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 2
- 238000003860 storage Methods 0.000 description 11
- 238000004519 manufacturing process Methods 0.000 description 9
- 238000011161 development Methods 0.000 description 5
- 230000006872 improvement Effects 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000002950 deficient Effects 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 235000020061 kirsch Nutrition 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
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
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- 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
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/20048—Transform domain processing
- G06T2207/20064—Wavelet transform [DWT]
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The embodiment of the invention provides a strip steel surface defect detection method, which comprises the following steps: acquiring a strip steel surface image; performing defect detection on the image by adopting a rapid classification detection method, judging whether the strip steel surface image contains defects, if so, performing accurate detection and classification on the strip steel surface image containing the defects by adopting a pyramid structure defect detection method, and determining the positions and types of the defects contained in the strip steel surface image; if not, carrying out defect detection on the strip steel surface image by adopting an image processing method, determining whether the strip steel surface image containing the defects contains the defects, if so, carrying out accurate detection and classification on the strip steel surface image containing the defects by adopting a defect detection method with a pyramid structure, determining the positions and types of the defects contained in the strip steel surface image, and if not, outputting the defects; according to the technical scheme of the invention, the real-time and accurate detection effect can be achieved, and the requirements of real-time performance and defect detection accuracy are met.
Description
Technical Field
The invention relates to the technical field of target detection, in particular to a method and a device for detecting surface defects of strip steel.
Background
With the rapid development of the economy in China, the steel industry is used as a supporting industry of manufacturing industry, the rapid development is carried out in recent years, and the strip steel is used as one of important products of the steel industry and is applied to various fields of industrial production and manufacturing, aerospace and the like. The surface quality of the strip steel seriously determines the final product performance, and the surface defect is an important factor influencing the surface quality of the strip steel, so that the detection of the surface defect of the strip steel and the improvement of the surface quality of the strip steel have important practical significance.
Conventionally, the detection of the surface defects of the strip steel mainly adopts a manual detection mode, and whether the surface contains defects is observed through human eyes. Due to the continuous improvement of the production speed, the manual detection method cannot reliably capture defect information, and the detection personnel are influenced by light, environment and physical factors to evaluate the detection results with different standards. The manual detection method is gradually difficult to adapt to the development requirement of the modern steel industry, most of factories can erect an area array camera on a production line at present, the shot strip steel picture is subjected to defect detection, the conventional defect detection method is to process the shot picture, the detection and the positioning of the strip steel surface defect are achieved according to the optimization and the improvement of the image processing method, but the defect detection accuracy is low, and the defect type cannot be accurately positioned and identified.
In recent years, the rapid development of the Internet, big data, scientific technology and the like, and an online detection system based on machine vision has been successfully applied to a steel strip production line, so that the efficient detection of the surface defects of the steel strip can be realized. The detection device of the type relies on a high-speed camera to shoot images of materials on a production line, and a computer is used for analyzing whether defects exist and what defects exist. The existing equipment generally uses a traditional image processing algorithm to detect defects, such as Fourier transform and the like, but due to the variety of defects, the actual defect detection accuracy is low, and manual secondary re-detection is needed. With the rapid development of the deep learning detection algorithm, the detection accuracy is greatly improved, but due to the limitation of equipment, the integration of the deep learning detection algorithm on a production line is difficult to detect defects of strip steel on the production line running at high speed in real time.
In summary, no detection method for detecting the surface defects of the strip steel can realize real-time and accurate detection.
Disclosure of Invention
The invention aims to provide a strip steel surface defect detection method, so as to achieve a real-time and accurate detection effect, and simultaneously meet the requirements of real-time performance and defect detection accuracy.
In order to achieve the above object, in one aspect, an embodiment of the present invention provides a method for detecting a surface defect of a strip steel, the method comprising:
acquiring a strip steel surface image;
performing defect detection on the strip steel surface image by adopting a rapid classification detection method, judging whether the strip steel surface image contains defects, if so, performing accurate detection and classification on the strip steel surface image containing the defects by adopting a pyramid structure defect detection method, and determining the positions and types of the defects contained in the strip steel surface image;
if not, carrying out defect detection on the strip steel surface image by adopting an image processing method, determining whether the strip steel surface image contains defects, if so, carrying out accurate detection and classification on the strip steel surface image containing the defects by adopting a defect detection method with a pyramid structure, determining the position and the type of the defects contained in the strip steel surface image, and if not, outputting the defects.
In another aspect, an embodiment of the present invention provides a device for detecting a surface defect of a strip steel, where the device includes:
the image acquisition unit is used for acquiring a strip steel surface image;
the first detection unit is used for detecting defects of the strip steel surface image by adopting a rapid classification detection method and judging whether the strip steel surface image contains defects or not;
the second detection unit is used for carrying out defect detection on the strip steel surface image, the defects of which are not detected by the first detection unit, by adopting an image processing method, and determining whether the strip steel surface image contains the defects or not;
and the third detection unit is used for accurately detecting and classifying the strip steel surface images containing the defects determined by the first detection unit and the second detection unit by adopting a defect detection method with a pyramid structure, and determining the positions and types of the defects contained in the strip steel surface images.
The technical scheme has the following beneficial effects: the two-stage detection method for the defects on the surface of the strip steel can rapidly and accurately detect the positions and the types of the defect information on the surface of the strip steel. By the first stage of classification detection, most images without defects can be rapidly filtered; the position and the category of the defect information on the surface of the strip steel can be accurately detected by the defect detection method of the pyramid structure at the second stage.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for detecting surface defects of strip steel according to an embodiment of the invention;
FIG. 2 is a schematic diagram of a device for detecting surface defects of a strip steel according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a flowchart of a method for detecting a surface defect of a strip steel according to an embodiment of the present invention is shown, and in order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings and examples.
The invention is based on the following experience rules found in actual production, taking the detection of the whole strip steel with the length of about 100 meters as an example, dividing the strip steel into a plurality of non-overlapped rectangular images with the width and the height of 10 cm, wherein more than half of the rectangular images are smooth surfaces without any defects and suspicious defects, the rest of the rectangular images are defect-free surfaces with unsmooth textures, and the rest of the rectangular images with defects. Based on the finding, the invention does not directly detect the defects of all images, but classifies the images first and then detects the defects, and because most images do not contain defects, the invention can reduce the number of images for detecting the defects, lighten the operation load of a system and rapidly finish the defect detection of the surface of the strip steel.
A strip steel surface defect detection method comprises the following steps:
s101, acquiring a shot strip steel surface image;
preferably, for the strip steel to be detected, an image of the surface of the strip steel is taken by an image acquisition device, and the surface information of the strip steel is recorded by the image, for example, the surface image of the strip steel can be obtained by using an industrial camera.
S102, performing defect detection on the strip steel surface image by adopting a rapid classification detection method, judging whether the strip steel surface image contains defects, if so, performing accurate detection and classification on the strip steel surface image containing the defects by adopting a pyramid structure defect detection method, and determining the positions and types of the defects contained in the strip steel surface image;
preferably, the rapid two-classification detection method comprises: performing edge detection on the strip steel surface image, binarizing the edge image of the strip steel surface image according to a preset edge pixel value threshold value, judging whether the strip steel surface image contains sharp defects according to the binarized edge image, if so, performing accurate detection and classification on the strip steel surface image by adopting a pyramid-structure defect detection method, and if not, performing defect detection on the strip steel surface image by adopting an image processing method.
In this embodiment, a specific description is given of a set of images established in advance. Taking as an example a set of images containing a certain number of defects X and a set of normal images Y, wherein Y comprises two subsets [ Y1, Y2], defined as: y1 is a normal image subset containing only smooth textures and Y2 is a normal image subset containing matte textures.
Further, performing edge detection on the strip steel surface image includes: an edge operator is used for extracting an edge image, and the edge image is binarized according to a preset edge pixel value threshold, the edge image binarization processing is that the pixel value of a pixel point of the edge image of the strip steel surface image, the pixel value of which is lower than the preset edge pixel value threshold, is set to 0, the pixel value of a pixel point of which the pixel value is higher than the preset edge pixel value threshold is set to 255, a Canny operator can be used for extracting the edge image, and the parameters of the Canny operator and the binarization pixel value threshold are set according to a known image set X, Y, and the method is as follows: under the precondition that the number of images in the set Y is misjudged to be lower than a certain proportion, the Canny operator parameter and the binary pixel value threshold value are adjusted, so that the images in the X are correctly judged to be the maximum number of the images containing the defects, so that the images containing the defects can be detected as much as possible under the condition of being as fast as possible, the detection speed is improved, whether the strip steel surface images to be detected contain sharp defects or not is judged according to the binary edge images, if so, the acquired strip steel surface images are accurately detected and classified by adopting a deep learning algorithm, and the defect detection is carried out on the images containing the defects which can not be identified yet after the step is defined by adopting an image processing method.
Further, the edge detection of the strip steel surface image further comprises the following operators: for example, the Roberts operator adopts the difference between two pixels adjacent in the diagonal direction, is an operator for searching edges by utilizing a local difference operator, and only uses the 2 x 2 neighborhood of the current pixel, and has the main defect of high sensitivity to noise because few pixels are used for approximating gradients; the Sobel operator firstly performs weighted average and then differentiation, then obtains gradient, and performs convolution operation by using the operator, wherein the maximum value of the two convolution kernels is used as an output bit of the point; the Prewitt operator is the same as the Sobel operator, and each point in the image is convolved by the two kernels, and the maximum value is taken as output; and (3) a Kirsch operator, wherein each point in the Kirsch operator image is convolved by 8 masks, each mask has the largest response to a specific edge direction, and the maximum value in all 8 directions is output as an edge amplitude image.
S103, if not, carrying out defect detection on the strip steel surface image by adopting an image processing method, determining whether the strip steel surface image contains defects, if so, carrying out accurate detection and classification on the strip steel surface image containing the defects by adopting a defect detection method of a pyramid structure, determining the position and the type of the defects contained in the strip steel surface image, and if not, outputting the defects.
Preferably, the image denoising and enhancing treatment is performed on the strip steel surface image, so that defect information can be reserved, background noise is removed, meanwhile, the denoised image defect information is enhanced, high-frequency characteristics of a defect area can be effectively amplified, the image defect information can be conveniently extracted, and the image denoising is performed specifically through a non-local mean method, so that the high-frequency characteristics in smooth texture are removed. The parameter determination method of the non-local mean method is as follows: according to the method, denoising is carried out on images in the training data set Y1 by a non-local mean value method, parameters are adjusted so that the Y1 set images reconstructed after filtering no longer contain smooth strip steel surface textures, high-frequency parts in normal strip steel surface textures can be effectively removed by the method, and frequency distribution of defect areas is obviously different from that of normal textures, so that the characteristics of the defect areas cannot be removed at the same time, and meanwhile, the high-frequency characteristics of the defect areas and the unsmooth normal texture areas are enhanced by using an adaptive histogram equalization method.
Preferably, the wavelet transform is performed on the image after noise reduction and enhancement, so that the wavelet transform can simultaneously process the low-frequency long-time characteristic and the high-frequency short-time characteristic of the signal, has good localization property, can effectively overcome the limitation of the Fourier transform in processing the non-stationary complex signal, has extremely strong self-adaptability, is convenient for extracting the frequency domain information of the image defect after the wavelet transform, is used for extracting the high-frequency information of the image of the defect area, simultaneously performs the high-pass filtering processing on the image after the wavelet transform, performs the wavelet transform processing on all the images without detecting the defect, and accurately detects and classifies all the processed images by adopting the defect detection method of a pyramid structure, thereby improving the detection accuracy and avoiding the occurrence of the omission detection event.
Further, the filtered image is binarized according to a preset image pixel value threshold, and 255 is set for the filtered image, that is, the pixel value is greater than the preset image pixel value threshold, and 0 is set for the filtered image, that is, the pixel value is less than the preset image pixel value threshold.
Further, counting the number of 255 pixels in the binarized image, wherein the number of 255 pixels in the binarized image is equal to the area of the defect area, if the area of the defect area is larger than a preset area value threshold, judging that the image contains the defect, and if the area of the defect area is smaller than the preset area value threshold, judging that the image does not contain the defect.
The parameter selection method comprises the following three steps: setting proper wavelet transformation, filter, binarization threshold and defect area threshold parameter selection range, and searching for reducing the number of judging normal pictures as defective pictures as far as possible on the premise that the number of missed judging defective pictures is 0 by a parameter grid searching method.
Preferably, the defect detection method of the pyramid structure includes:
the method is characterized in that a characteristic network with smaller and smaller size is gradually generated after the strip steel surface image passes through a depth convolution network, a target area with smaller area can be detected by a characteristic image with shallower depth, namely larger size, a target area with larger area can be detected by a characteristic image with deeper depth, namely smaller size, and the detection process is defined as pyramid structure defect detection.
Extracting the area size of all defects in a certain number of pre-established defect image sets X and forming an area set M, squaring each element in the M to obtain a rectangular side length set L, clustering the set L by using a K-means clustering (K-means) method, traversing the number of classes upwards from 2 to enable the K-means clustering loss to meet the requirement, namely stopping, the number of classes, namely the number of pyramid layers, and enabling the clustering center to be the size of a pyramid priori detection frame of each layer;
generating a characteristic map of the strip steel surface image in each layer of pyramid, performing candidate region proposal in each layer of pyramid characteristic map, and after removing overlapped candidate regions by using a non-maximum suppression method, proposing n positions which are possibly defect regions for each layer of pyramid.
And (3) putting the n-pyramid-layer number defect candidate areas into a convolution network, and outputting the defect possibility, defect type and fine-tuned area position of each candidate area.
In another aspect, an embodiment of the present invention provides a device for detecting a surface defect of a strip steel, where the device includes:
an image acquisition unit 21 for acquiring an image of the surface of the strip steel;
a first detecting unit 22, configured to detect a defect of the strip steel surface image by using a rapid classification detecting method, and determine whether the strip steel surface image includes a defect;
a second detecting unit 23, configured to detect a defect in the strip surface image that is not detected by the first detecting unit by using an image processing method, and determine whether the strip surface image contains the defect;
and the third detection unit 24 is used for accurately detecting and classifying the strip steel surface images containing the defects determined by the first detection unit and the second detection unit by adopting a defect detection method with a pyramid structure, and determining the positions and types of the defects contained in the strip steel surface images.
Preferably, the first detecting unit 22 includes:
and carrying out edge detection on the strip steel surface image, binarizing the edge image of the strip steel surface image according to a preset edge pixel value threshold value, and judging whether the strip steel surface image contains sharp defects according to the binarized edge image.
Preferably, the second detecting unit 23 includes:
carrying out image noise reduction treatment on the strip steel surface image;
performing image enhancement processing on the noise-reduced strip steel surface image;
performing wavelet transformation processing on the enhanced image, extracting frequency domain information of image defects after the wavelet transformation processing, and performing high-pass filtering processing on the image after the wavelet transformation processing;
binarizing the filtered image according to a preset image pixel value threshold;
and counting the area of the defect area in the binarized image, judging the image to contain the defect if the area is larger than a preset area value threshold value, and judging the image to certainly not contain the defect if the area is smaller than the preset area value threshold value.
Preferably, the third detecting unit 24 further includes:
the number of layers of the pyramid is determined, the area of the defect in the pre-established defect image set X is extracted to form an area set M, each element in the M is squared to obtain a rectangular side length set L, and clustering processing is carried out on the set L, wherein the number of classes is the number of layers of the pyramid.
It should be understood that the specific order or hierarchy of steps in the processes disclosed are examples of exemplary approaches. Based on design preferences, it is understood that the specific order or hierarchy of steps in the processes may be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in a sample order, and are not meant to be limited to the specific order or hierarchy presented.
In the foregoing detailed description, various features are grouped together in a single embodiment for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments of the subject matter require a relatively clear and comprehensive understanding of the meaning or content of the features recited in each claim. Rather, as the following claims reflect, invention lies in less than all features of a single disclosed embodiment. Thus the following claims are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate preferred embodiment of this invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. As will be apparent to those skilled in the art; various modifications to these embodiments will be readily apparent, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the disclosure. Thus, the present disclosure is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description includes examples of one or more embodiments. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the aforementioned embodiments, but one of ordinary skill in the art may recognize that many further combinations and permutations of various embodiments are possible. Accordingly, the embodiments described herein are intended to embrace all such alterations, modifications and variations that fall within the scope of the appended claims. Furthermore, as used in the specification or claims, the term "comprising" is intended to be inclusive in a manner similar to the term "comprising," as interpreted when employed as a transitional word in a claim. Furthermore, any use of the term "or" in the specification of the claims is intended to mean "non-exclusive or".
Those of skill in the art will further appreciate that the various illustrative logical blocks (illustrative logical block), units, and steps described in connection with the embodiments of the invention may be implemented by electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components (illustrative components), elements, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design requirements of the overall system. Those skilled in the art may implement the described functionality in varying ways for each particular application, but such implementation is not to be understood as beyond the scope of the embodiments of the present invention.
The various illustrative logical blocks or units described in the embodiments of the invention may be implemented or performed with a general purpose processor, a digital signal processor, an Application Specific Integrated Circuit (ASIC), a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described. A general purpose processor may be a microprocessor, but in the alternative, the general purpose processor may be any fixed processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. In an example, a storage medium may be coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC, which may reside in a user terminal. In the alternative, the processor and the storage medium may reside as distinct components in a user terminal.
In one or more exemplary designs, the above-described functions of embodiments of the present invention may be implemented in hardware, software, firmware, or any combination of the three. If implemented in software, the functions may be stored on a computer-readable medium or transmitted as one or more instructions or code on the computer-readable medium. Computer readable media includes both computer storage media and communication media that facilitate transfer of computer programs from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer. For example, such computer-readable media may include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that may be used to carry or store program code in the form of instructions or data structures and other data structures that may be read by a general or special purpose computer, or a general or special purpose processor. Further, any connection is properly termed a computer-readable medium, e.g., if the software is transmitted from a website, server, or other remote source via a coaxial cable, fiber optic cable, twisted pair, digital Subscriber Line (DSL), or wireless such as infrared, radio, and microwave, and is also included in the definition of computer-readable medium. The disks (disks) and disks (disks) include compact disks, laser disks, optical disks, DVDs, floppy disks, and blu-ray discs where disks usually reproduce data magnetically, while disks usually reproduce data optically with lasers. Combinations of the above may also be included within the computer-readable media.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (4)
1. A method for detecting surface defects of a strip steel, comprising the steps of:
acquiring a strip steel surface image;
performing defect detection on the strip steel surface image by adopting a rapid classification detection method, judging whether the strip steel surface image contains sharp defects, if so, performing accurate detection and classification on the strip steel surface image containing the defects by adopting a pyramid structure defect detection method, and determining the positions and types of the defects contained in the strip steel surface image;
if the defect is not included, performing defect detection on the strip steel surface image by adopting an image processing method, determining whether the strip steel surface image includes the defect, if so, performing accurate detection and classification on the strip steel surface image including the defect by adopting a defect detection method of a pyramid structure, determining the position and the type of the defect included in the strip steel surface image, and if not, outputting a defect-free image;
the rapid classification detection method comprises the following steps:
performing edge detection on the strip steel surface image, binarizing the edge image of the strip steel surface image according to a preset edge pixel value threshold value, and judging whether the strip steel surface image contains sharp defects according to the binarized edge image;
the image processing method comprises the following steps:
carrying out image noise reduction treatment on the strip steel surface image;
performing image enhancement processing on the noise-reduced strip steel surface image;
performing wavelet transformation processing on the enhanced image, extracting frequency domain information of image defects after the wavelet transformation processing, and performing high-pass filtering processing on the image after the wavelet transformation processing;
binarizing the filtered image according to a preset image pixel value threshold;
counting the area of a defect area in the binarized image, judging the image to contain the defect if the area is larger than a preset area value threshold value, and judging the image to certainly not contain the defect if the area is smaller than the preset area value threshold value;
the method for precisely detecting and classifying the band steel surface image containing the defects by adopting the defect detection method with a pyramid structure, and determining the position and the type of the defects contained in the band steel surface image further comprises the following steps:
the number of layers of the pyramid is determined, the area of the defect in the pre-established defect image set X is extracted to form an area set M, each element in the M is squared to obtain a rectangular side length set L, and clustering processing is carried out on the set L, wherein the number of classes is the number of layers of the pyramid.
2. The method for detecting defects on a surface of a strip steel according to claim 1, wherein the method for precisely detecting and classifying defects in the surface image of the strip steel containing defects by using the defect detection method of a pyramid structure, and determining the positions and types of the defects contained in the surface image of the strip steel comprises:
generating a characteristic map of the strip steel surface image in each layer of pyramid, performing candidate region proposal in each layer of pyramid characteristic map, and after removing overlapped candidate regions by using a non-maximum suppression method, proposing n positions which are possibly defect regions for each layer of pyramid;
and (3) putting the n-pyramid-layer number defect candidate areas into a convolution network, and outputting the defect possibility, defect type and fine-tuned area position of each candidate area.
3. The method for detecting surface defects of a strip steel according to claim 1, wherein the binarizing process comprises:
the binarization processing is to set the pixel value of the pixel point of which the pixel value is lower than the preset pixel value threshold value of the strip steel surface image to 0 and set the pixel value of the pixel point of which the pixel value is higher than the preset pixel value threshold value to 255.
4. A strip steel surface defect detection device, characterized by comprising:
the image acquisition unit is used for acquiring a strip steel surface image;
the first detection unit is used for detecting defects of the strip steel surface image by adopting a rapid classification detection method and judging whether the strip steel surface image contains sharp defects or not;
the second detection unit is used for carrying out defect detection on the strip steel surface image, the sharp defects of which are not detected by the first detection unit, by adopting an image processing method, and determining whether the strip steel surface image contains defects or not;
the third detection unit is used for accurately detecting and classifying the strip steel surface images containing the defects determined by the first detection unit and the second detection unit by adopting a defect detection method with a pyramid structure, and determining the positions and types of the defects contained in the strip steel surface images;
the first detection unit includes:
performing edge detection on the strip steel surface image, binarizing the edge image of the strip steel surface image according to a preset edge pixel value threshold value, and judging whether the strip steel surface image contains sharp defects according to the binarized edge image;
the second detection unit includes:
carrying out image noise reduction treatment on the strip steel surface image;
performing image enhancement processing on the noise-reduced strip steel surface image;
performing wavelet transformation processing on the enhanced image, extracting frequency domain information of image defects after the wavelet transformation processing, and performing high-pass filtering processing on the image after the wavelet transformation processing;
binarizing the filtered image according to a preset image pixel value threshold;
counting the area of a defect area in the binarized image, judging the image to contain the defect if the area is larger than a preset area value threshold value, and judging the image to certainly not contain the defect if the area is smaller than the preset area value threshold value;
the third detection unit further includes:
the number of layers of the pyramid is determined, the area of the defect in the pre-established defect image set X is extracted to form an area set M, each element in the M is squared to obtain a rectangular side length set L, and clustering processing is carried out on the set L, wherein the number of classes is the number of layers of the pyramid.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911107307.0A CN111047556B (en) | 2019-11-13 | 2019-11-13 | Strip steel surface defect detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911107307.0A CN111047556B (en) | 2019-11-13 | 2019-11-13 | Strip steel surface defect detection method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111047556A CN111047556A (en) | 2020-04-21 |
| CN111047556B true CN111047556B (en) | 2024-04-05 |
Family
ID=70232426
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911107307.0A Active CN111047556B (en) | 2019-11-13 | 2019-11-13 | Strip steel surface defect detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111047556B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112785612B (en) * | 2020-08-28 | 2022-09-13 | 青岛经济技术开发区海尔热水器有限公司 | Image edge detection method based on wavelet transformation |
| WO2022160167A1 (en) * | 2021-01-28 | 2022-08-04 | 东莞职业技术学院 | Strip steel surface defect detection method based on convolutional neural network model training |
| CN113723219A (en) * | 2021-08-10 | 2021-11-30 | 浙江大华技术股份有限公司 | Method for identifying breakage of insulating member and related device |
| CN114199879B (en) * | 2021-11-23 | 2024-06-18 | 北京科技大学 | Identification method for aggregated defects on surface of cold-rolled strip steel |
| CN114454137A (en) * | 2022-04-12 | 2022-05-10 | 西南交通大学 | Steel structure damage intelligent inspection method and system based on binocular vision and robot |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794491A (en) * | 2015-04-28 | 2015-07-22 | 重庆大学 | Fuzzy clustering steel plate surface defect detection method based on pre classification |
| CN108230317A (en) * | 2018-01-09 | 2018-06-29 | 北京百度网讯科技有限公司 | Steel plate defect detection sorting technique, device, equipment and computer-readable medium |
| CN109829900A (en) * | 2019-01-18 | 2019-05-31 | 创新奇智(北京)科技有限公司 | A kind of steel coil end-face defect inspection method based on deep learning |
| CN109859163A (en) * | 2018-12-19 | 2019-06-07 | 重庆邮电大学 | A kind of LCD defect inspection method based on feature pyramid convolutional neural networks |
| CN110189290A (en) * | 2019-04-08 | 2019-08-30 | 广东工业大学 | Metal surface fine defects detection method and device based on deep learning |
| CN110223296A (en) * | 2019-07-08 | 2019-09-10 | 山东建筑大学 | A kind of screw-thread steel detection method of surface flaw and system based on machine vision |
-
2019
- 2019-11-13 CN CN201911107307.0A patent/CN111047556B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794491A (en) * | 2015-04-28 | 2015-07-22 | 重庆大学 | Fuzzy clustering steel plate surface defect detection method based on pre classification |
| CN108230317A (en) * | 2018-01-09 | 2018-06-29 | 北京百度网讯科技有限公司 | Steel plate defect detection sorting technique, device, equipment and computer-readable medium |
| CN109859163A (en) * | 2018-12-19 | 2019-06-07 | 重庆邮电大学 | A kind of LCD defect inspection method based on feature pyramid convolutional neural networks |
| CN109829900A (en) * | 2019-01-18 | 2019-05-31 | 创新奇智(北京)科技有限公司 | A kind of steel coil end-face defect inspection method based on deep learning |
| CN110189290A (en) * | 2019-04-08 | 2019-08-30 | 广东工业大学 | Metal surface fine defects detection method and device based on deep learning |
| CN110223296A (en) * | 2019-07-08 | 2019-09-10 | 山东建筑大学 | A kind of screw-thread steel detection method of surface flaw and system based on machine vision |
Non-Patent Citations (1)
| Title |
|---|
| 基于图像预处理的神经网络带钢缺陷检测;管声启等;《钢铁研究》;第41卷(第01期);第22-26页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111047556A (en) | 2020-04-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111047556B (en) | Strip steel surface defect detection method and device | |
| WO2022042579A1 (en) | Lcd screen defect detection method and apparatus | |
| CN109410230B (en) | Improved Canny image edge detection method capable of resisting noise | |
| CN109507192B (en) | Magnetic core surface defect detection method based on machine vision | |
| WO2021109697A1 (en) | Character segmentation method and apparatus, and computer-readable storage medium | |
| CN102739952B (en) | Multi-view image capturing method and application system thereof | |
| CN109584198B (en) | Method and device for evaluating quality of face image and computer readable storage medium | |
| CN110378902B (en) | Scratch detection method under high noise background | |
| CN107240086B (en) | A kind of fabric defects detection method based on integral nomography | |
| CN116542982B (en) | Departure judgment device defect detection method and device based on machine vision | |
| CN117437219B (en) | Magnetic glue inductance irregular appearance defect detection method based on computer vision | |
| CN116109637B (en) | System and method for detecting appearance defects of turbocharger impeller based on vision | |
| CN115619793A (en) | Power adapter appearance quality detection method based on computer vision | |
| CN115359044A (en) | Metal part surface scratch detection method based on image enhancement | |
| CN115619775B (en) | Material counting method and device based on image recognition | |
| CN106023160B (en) | Blast furnace charge level method for detecting image edge and device | |
| CN112669295A (en) | Lithium battery pole piece defect detection method based on secondary threshold segmentation theory | |
| CN110007493A (en) | Broken bright spot detection method in liquid crystal display | |
| CN115587966A (en) | Method and system for detecting whether parts are missing or not under condition of uneven illumination | |
| CN117764989B (en) | Visual-aided display screen defect detection method | |
| CN116934746B (en) | Scratch defect detection method, system, equipment and medium thereof | |
| CN117011291B (en) | Watch shell quality visual detection method | |
| CN115994870B (en) | Image processing method for enhancing denoising | |
| CN116524269A (en) | Visual recognition detection system | |
| CN115496984A (en) | Ceramic tile finished product defect automatic identification method and device, intelligent terminal and storage medium |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhang Guanghao Inventor after: Cui Dongshun Inventor after: Wang Shuigen Inventor after: Qian Xing Inventor before: Zhang Guanghao Inventor before: Cui Dongshun Inventor before: Wang Shuigen Inventor before: Huang Guangbin Inventor before: Qian Xing |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |