CN113177949B - Large-size rock particle feature recognition method and device - Google Patents
Large-size rock particle feature recognition method and device Download PDFInfo
- Publication number
- CN113177949B CN113177949B CN202110410018.9A CN202110410018A CN113177949B CN 113177949 B CN113177949 B CN 113177949B CN 202110410018 A CN202110410018 A CN 202110410018A CN 113177949 B CN113177949 B CN 113177949B
- Authority
- CN
- China
- Prior art keywords
- rock
- image
- size
- data
- image data
- 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
- 239000011435 rock Substances 0.000 title claims abstract description 148
- 239000002245 particle Substances 0.000 title claims abstract description 96
- 238000000034 method Methods 0.000 title claims abstract description 62
- 230000011218 segmentation Effects 0.000 claims abstract description 49
- 238000003709 image segmentation Methods 0.000 claims description 29
- 238000004458 analytical method Methods 0.000 claims description 20
- 238000012937 correction Methods 0.000 claims description 20
- 230000007613 environmental effect Effects 0.000 claims description 16
- 238000013528 artificial neural network Methods 0.000 claims description 15
- 238000005259 measurement Methods 0.000 claims description 15
- 238000001514 detection method Methods 0.000 claims description 13
- 230000000877 morphologic effect Effects 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 12
- 238000003384 imaging method Methods 0.000 claims description 11
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 9
- 229910052802 copper Inorganic materials 0.000 claims description 9
- 239000010949 copper Substances 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 230000003044 adaptive effect Effects 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 8
- 238000013135 deep learning Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000011282 treatment Methods 0.000 claims description 6
- 230000001186 cumulative effect Effects 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000012706 support-vector machine Methods 0.000 claims description 4
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 238000012512 characterization method Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 238000005315 distribution function Methods 0.000 claims description 3
- 230000003628 erosive effect Effects 0.000 claims description 3
- 238000003711 image thresholding Methods 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 238000013145 classification model Methods 0.000 claims description 2
- 239000003086 colorant Substances 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 238000011065 in-situ storage Methods 0.000 claims description 2
- 230000007246 mechanism Effects 0.000 claims description 2
- 238000011176 pooling Methods 0.000 claims description 2
- 238000002360 preparation method Methods 0.000 claims description 2
- 230000001629 suppression Effects 0.000 claims description 2
- 238000013139 quantization Methods 0.000 claims 1
- 238000003703 image analysis method Methods 0.000 abstract description 6
- 238000000691 measurement method Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 description 7
- 239000004575 stone Substances 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000010339 dilation Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000011179 visual inspection Methods 0.000 description 2
- 235000019738 Limestone Nutrition 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000010438 granite Substances 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 239000006028 limestone Substances 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000005445 natural material Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000005303 weighing 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
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06T5/90—Dynamic range modification of images or parts thereof
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color 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
- G06T2207/20036—Morphological image processing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a large-size rock particle feature recognition method and device. Wherein the method comprises the following steps: acquiring rock image data; generating a preset segmentation algorithm according to the rock image data; identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics; and outputting the geometric parameter characteristics of the rock. The invention solves the technical problems that the manual measurement and image analysis method in the prior art is not accurate and efficient for analyzing the large-size rock.
Description
Technical Field
The invention relates to the field of image recognition, in particular to a large-size rock particle feature recognition method and device.
Background
The rock particles with larger size have various lithologies, usually granite or limestone, and also have part of concrete broken stones generated in the building and paving and dismantling processes, and are widely applied to geotechnical engineering and hydraulic engineering, and are mainly used for protecting coastlines, riverbeds, bridge decks, pile foundations and other coastline structures. As a natural material, the reliability and sustainable usability of rockfill materials requires quality control throughout the design, production, transportation, installation, inspection and maintenance stages, while the different quarrying processes and mineral lithology bring randomness to the aggregate quality produced. The traditional means require the weight of each rock to be measured one by one, which is time-consuming and labor-consuming, whereas standard screening methods cannot be used to obtain the geometric size and shape characteristics of such larger-particle-size piles, the main methods at present being based on visual or manual measurements. Thus, there is a need for reliable imaging techniques to easily and quickly process the stone image for further analysis. Currently, imaging-based analysis techniques have been widely developed and the size and geometry characteristics of aggregate particles are characterized by image analysis techniques. Image-based indoor analysis methods have focused mainly on shape analysis using cameras attached to table-sized devices on which a set of prepared particles are placed for photo acquisition. For example, the collective imaging system (AIMS) uses a device composed of one camera and two light sources; an integrated image analyzer at the university of illinois (E-uiia) relies on cameras mounted on three orthogonal axes to take three views of a three-dimensional particle. The three-dimensional particles produced by these methods may lack shape details, such as local concave curvature on the surface; furthermore, these devices are often inconvenient to carry and are cumbersome to measure on site. Laboratory imaging equipment is mainly used for rock shape acquisition, whereas rock geometry analysis systems such as wilrag have been developed to obtain particle size distribution in the field, but have limited effectiveness due to shadows of particles in the image and mutual pile-up overlapping affecting the quality and accuracy of the image. Thus, image-based measurement of aggregate particle close-packed (or loose-packed) volumes requires image segmentation of the volumes to determine the size and morphology characteristics of individual particles. To achieve this goal, there is an urgent need to develop efficient computer vision algorithms based on conventional computer vision techniques and/or emerging deep learning techniques to quickly obtain the geometric and shape characteristics of large-size rock mass aggregates.
In the prior art, two ways are generally adopted to identify large-size rock characteristics:
1. manual measurement: for the measurement of large-size rock particle geometry and property characteristics, visual inspection and manual measurement are currently relied on, and visual inspection is largely dependent on experience and expertise of practitioners. The manual measurement method generally needs to use some measurement tools or key particles and sample particles as references to assist judgment, and the mass acquisition can adopt a method of directly weighing single particles, or can also adopt a method of performing size-mass conversion after measuring the particle size, and currently, the method of measuring the middle size or circumference from three orthogonal axes and estimating the volume based on cuboid assumption or average sphere cube is generally adopted, so that the mass is estimated. However, the rough estimates provided by visual and manual measurements do not necessarily represent actual rock particle characteristics.
2. The image analysis method Wipfrag software is generally used for determining the particle size of crushed stone particles, and finally directly outputting a grading curve, so that the software can obtain a result with lower accuracy. Aggregate imaging system (AIMS) consists of a camera mounted on a slide cover and two light sources, measuring a particle size of 15mm at maximum. The camera in the aggregate image analyzer of the university of illinois (E-UIAIA) is mounted along three orthogonal axes for obtaining three views (front, top and side) of three-dimensional particles, and a set of orthogonal views of each particle is obtained by post-processing, with a maximum particle size of up to 7.6 cm. And the aggregate imaging system of the 3D laser scans single particles and analyzes the generated 3D grid model, and the maximum scannable sample particle size is 60m. ImageJ can perform image segmentation and then calculate the particle size of each particle using Excel, but the accuracy of the results is also low, mainly due to large errors in image segmentation. Three-dimensional solid models constructed based on X-ray CT images are used for aggregate shape characterization and volume estimation. The design of these systems is mostly laboratory dependent, they may not be easily transported, assembled and deployed into field applications, especially involving advanced equipment such as 3D laser scanners or X-ray CT scanners, etc. Furthermore, most systems have a maximum particle size limitation, which limits their application in the treatment of large-size rock particles. The lighting conditions of these systems are controlled by using a backlight or a plurality of light sources, with the aim of minimizing shadows and reflection effects. Furthermore, image filtering algorithms originally developed for laboratory conditions may also present difficulties in terms of accuracy under field lighting conditions. Thus, laboratory imaging systems are not directly suitable for field inspection.
In view of the above problems, no effective solution has been proposed at present.
Disclosure of Invention
The embodiment of the invention provides a large-size rock particle feature recognition method and device, which are used for solving the technical problems that in the prior art, manual measurement and image analysis methods are not accurate enough and efficient in analyzing large-size rock.
According to an aspect of an embodiment of the present invention, there is provided a large-size rock particle feature recognition method, including: acquiring rock image data; generating a preset segmentation algorithm according to the rock image data; identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics; and outputting the geometric parameter characteristics of the rock.
Optionally, after the acquiring rock image data, the method further comprises: environmental data is acquired.
Optionally, the rock image data is image data acquired by a plurality of image acquisition terminals.
Optionally, after the outputting the rock geometric parameter feature, the method further comprises: and integrating the preset segmentation algorithm to an application end with a user interface.
According to another aspect of the embodiment of the present invention, there is also provided a large-sized rock particle feature recognition apparatus, including: the acquisition module is used for acquiring rock image data; the generation module is used for generating a preset segmentation algorithm according to the rock image data; the identification module is used for identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics; and the output module is used for outputting the rock geometric parameter characteristics.
Optionally, the apparatus further includes: the acquisition module is also used for acquiring the environment data.
Optionally, the rock image data is image data acquired by a plurality of image acquisition terminals.
Optionally, the apparatus further includes: and the integration module is used for integrating the preset segmentation algorithm to an application end with a user interface.
According to another aspect of the embodiment of the present invention, there is also provided a nonvolatile storage medium, where the nonvolatile storage medium includes a stored program, and when the program runs, the program controls a device in which the nonvolatile storage medium is located to execute a large-size rock particle feature identification method.
According to another aspect of the embodiment of the present invention, there is also provided an electronic device including a processor and a memory; the memory stores computer readable instructions, and the processor is configured to execute the computer readable instructions, where the computer readable instructions execute a large-size rock particle feature recognition method when executed.
In the embodiment of the invention, rock image data are acquired; generating a preset segmentation algorithm according to the rock image data; identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics; the method for outputting the geometric parameter characteristics of the rock solves the technical problems that in the prior art, the manual measurement and the image analysis method are not accurate and efficient for analyzing the large-size rock.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a flow chart of a method for identifying characteristics of large-size rock particles according to an embodiment of the application;
FIG. 2 is a block diagram of a large-size rock particle signature recognition device in accordance with an embodiment of the present application;
fig. 3 is a schematic view of a photographing apparatus according to an embodiment of the present application;
FIG. 4 is a graph of segmentation effects according to an embodiment of the application;
wherein fig. 4 (a) separates particles, (b) contacts or overlaps particles, (c) densely stacks the original image of particles. (d) Separating particles, (e) contacting or overlapping particles, (f) densely stacking particle segmented images of particles. (g) Separating particles, (h) contacting or overlapping particles, (i) densely stacking particle segmented images of particles.
Detailed Description
In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an embodiment of the present invention, there is provided a method embodiment of a large-size rock particle feature recognition method, it being noted that the steps shown in the flowchart of the drawings may be performed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in an order different from that shown or described herein.
Example 1
Fig. 1 is a flowchart of a large-sized rock particle feature recognition method according to an embodiment of the present invention, as shown in fig. 1, the method comprising the steps of:
step S102, rock image data is acquired.
Specifically, for a single particle rock sample, in order to analyze large-size rock data, the embodiment of the invention needs to acquire corresponding large-size rock sample image data first, and further processing, analysis and treatment are performed according to the acquired rock image data.
Optionally, after the acquiring rock image data, the method further comprises: environmental data is acquired.
Specifically, after the image data of the rock is obtained, the image recognition and analysis are more accurate by adding the environmental data factors, and the environmental data can also be obtained, wherein the environmental data comprises weather data, light data and the like, the environmental data can be obtained by collecting and outputting the surrounding environment through the image collecting equipment, and the environmental data such as the local weather at the time can also be determined through remote server data call.
Optionally, the rock image data is image data acquired by a plurality of image acquisition terminals.
Specifically, as shown in fig. 3, the image acquisition apparatus is composed of five main parts: three smart phones, high resolution cameras and remote shutter control, three copper pipes with the length of 1.52 meters are spliced by a joint for shorter pipelines, so that the copper pipes can be assembled and disassembled easily, the smart phones are more suitable for field use, a 10 kg patio umbrella base is used as an anchorage of the copper pipes, three blue curtains with the length of 1.52 meters multiplied by 1.52 meters are used as the background and the bottom surface, and three camera tripods are used for fixing the smart phones on the top/front/side surfaces. One tripod is provided with a cantilever, so that the top of the rock can be clearly photographed. After the equipment is debugged, rock photos with different sizes are taken from different observation angles to form a database. After each shot is completed, the particles are randomly rotated by an angle such that each particle is repeatedly shot at least three times. The purpose of the repeated rotation test is to check the suitability of the imaging procedure and to further study the changes that result from observing individual rocks from different angles. The accuracy of the size (and shape) measurements was verified by comparison with manually measuring the size and weight of individual polished rocks.
Step S104, a preset segmentation algorithm is generated according to the rock image data.
And step S106, identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics.
And (3) carrying out reconstruction quantification on the volume of the object by combining a multi-view information-based image segmentation algorithm with a three-dimensional reconstruction algorithm.
1) Image segmentation algorithm-color-based object detection image segmentation algorithm: the proposed image segmentation algorithm is a color-based object detection image segmentation algorithm that includes three parts, color space representation, foreground-background contrast enhancement, adaptive thresholding, and morphological denoising.
a) Color space representation—cie L a b color space: in the CIE L x a x b x space, L x channels represent luminance or intensity values, and a x channel and b x channel track the green to red, blue to yellow transition, respectively. The CIE L x a x b x space can well eliminate shadowing effects by separating useful information into its a x and b x channels. Depending on the object color and background color, useful object information may be accumulated in the a-channel, b-channel, or both.
b) Foreground-background contrast enhancement: first, obtaining pixel histograms of an a channel and a b channel, and obtaining a pixel cumulative distribution function through the pixel histograms.
The contrast enhancement of the foreground and the background is achieved by the color distance, the gray-scale intensity of the pixel represents the distance of the color referenced by the foreground representative color, the closer the pixel color is to the foreground representative color, the smaller the color distance in the distance map, the darker the intensity, and vice versa. The method helps to better contrast the background and foreground and further eliminate shadow effects.
c) Adaptive thresholding and morphological denoising
Based on the enhanced distance map, image thresholding (i.e., binarization) is applied to obtain a binarized image, and the threshold setting algorithm can follow either a fixed threshold or a flexible threshold, i.e., an adaptive threshold setting method. Because digital images are pixel-wise discretized, binary images typically contain a large number of noisy pixels, requiring denoising to remove the noisy pixels and discontinuities along the boundary of the object. A series of morphological operations are performed on the binarized image, including image erosion, dilation, hole filling, etc. An unrecognizable object, such as a device or field environment, is also removed, typically in an area that appears closer to the image boundary.
(2) Three-dimensional volume reconstruction algorithm-three-dimensional volume reconstruction algorithm based on orthogonal calibration and volume correction
a) Least square method orthogonal calibration
Orthogonal calibration is required due to photogrammetry errors. The normalized orthonormal dimension is obtained from the following set of linear equations:
Ax＝b
the linear system solution minimizes the residual term, i.e
b) Volume correction
The volume of the segmented object may be calculated as a "voxel", i.e. a three-dimensional cuboid pixel. The volume of the reconstructed body can be obtained by voxel ratio between the rock and the calibration sphere, but the volume of the reconstructed object is always larger than or equal to the volume of the actual object. To correct for the larger volume after reconstruction, a correction factor was used, which was taken to be 0.95 after extensive experimental summary.
c) Resolution correction
Since the segmentation algorithm is based on the front-background contrast, the detected boundary will be slightly smaller than the actual object boundary. This result can lead to resolution-based overestimation, which is controlled by two parameters. The first parameter is the relative size ratio of the rock to the calibration sphere, and the second parameter is the absolute pixel occupancy of the calibration sphere. The following correction coefficient equation is used:
after a large number of experiments, the value of the correction parameter is generally 0.90.
And step S108, outputting the rock geometric parameter characteristics.
Specifically, in the embodiment of the present invention, after the rock geometric parameter feature is calculated by the preset image segmentation algorithm, the result value expected by the embodiment of the present invention is obtained, and then the rock geometric parameter feature may be output to the display section or the storage end, so that the user may conveniently view and apply the rock geometric parameter feature.
Optionally, after the outputting the rock geometric parameter feature, the method further comprises: and integrating the preset segmentation algorithm to an application end with a user interface.
Specifically, in order to enable an application end with a user interface to obtain an accurate preset segmentation algorithm, the preset segmentation algorithm (algorithm for calculating geometric parameter characteristics of rock) displayed or output in the embodiment of the invention is integrated, that is, fused to a corresponding application end, so that large-size rock identification can be continued later.
Through the embodiment, the technical problems that the manual measurement and the image analysis method in the prior art are not accurate and efficient for analyzing the large-size rock are solved.
Example two
Fig. 2 is a block diagram of a large-sized rock particle characteristic recognition apparatus according to an embodiment of the present invention, as shown in fig. 2, the apparatus including:
An acquisition module 20 for acquiring rock image data.
Specifically, for a single particle rock sample, in order to analyze large-size rock data, the embodiment of the invention needs to acquire corresponding large-size rock sample image data first, and further processing, analysis and treatment are performed according to the acquired rock image data.
Optionally, after the acquiring rock image data, the method further comprises: environmental data is acquired.
Specifically, after the image data of the rock is obtained, in order to make the image recognition and analysis more accurate by adding the environmental data factor, environmental data can also be obtained, where the environmental data includes weather data, light data, and the like, and the environmental data can be obtained by collecting and outputting the surrounding environment through the image collecting device, or determining the environmental data such as the local weather at the time through a remote server data call.
Optionally, the rock image data is image data acquired by a plurality of image acquisition terminals.
Specifically, as shown in fig. 3, the image acquisition apparatus is composed of five main parts: three smart phones, high resolution cameras and remote shutter control, three copper pipes with the length of 1.52 meters are spliced by a joint for shorter pipelines, so that the copper pipes can be assembled and disassembled easily, the smart phones are more suitable for field use, a 10 kg patio umbrella base is used as an anchorage of the copper pipes, three blue curtains with the length of 1.52 meters multiplied by 1.52 meters are used as the background and the bottom surface, and three camera tripods are used for fixing the smart phones on the top/front/side surfaces. One tripod is provided with a cantilever, so that the top of the rock can be clearly photographed. After the equipment is debugged, rock photos with different sizes are taken from different observation angles to form a database. After each shot is completed, the particles are randomly rotated by an angle such that each particle is repeatedly shot at least three times. The purpose of the repeated rotation test is to check the suitability of the imaging procedure and to further study the changes that result from observing individual rocks from different angles. The accuracy of the size (and shape) measurements was verified by measuring the size and weight of individual polished rocks by hand.
And the generating module 22 is used for generating a preset segmentation algorithm according to the rock image data.
And the identification module 24 is used for identifying the rock image data according to the preset segmentation algorithm to obtain the rock geometric parameter characteristics.
And (3) carrying out reconstruction quantification on the volume of the object by combining a multi-view information-based image segmentation algorithm with a three-dimensional reconstruction algorithm.
1) Image segmentation algorithm-color-based object detection image segmentation algorithm: the proposed image segmentation algorithm is a color-based object detection image segmentation algorithm that includes three parts, color space representation, foreground-background contrast enhancement, adaptive thresholding, and morphological denoising.
a) Color space representation—cie L a b color space: in the CIE L x a x b x space, L x channels represent luminance or intensity values, and a x channel and b x channel track the green to red, blue to yellow transition, respectively. The CIE L x a x b x space can well eliminate shadowing effects by separating useful information into its a x and b x channels. Depending on the object color and background color, useful object information may be accumulated in the a-channel, b-channel, or both.
b) Foreground-background contrast enhancement: first, obtaining pixel histograms of an a channel and a b channel, and obtaining a pixel cumulative distribution function through the pixel histograms.
The contrast enhancement of the foreground and the background is achieved by the color distance, the gray-scale intensity of the pixel represents the distance of the color referenced by the foreground representative color, the closer the pixel color is to the foreground representative color, the smaller the color distance in the distance map, the darker the intensity, and vice versa. The method helps to better contrast the background and foreground and further eliminate shadow effects.
c) Adaptive thresholding and morphological denoising
Based on the enhanced distance map, image thresholding (i.e., binarization) is applied to obtain a binarized image, and the threshold setting algorithm can follow either a fixed threshold or a flexible threshold, i.e., an adaptive threshold setting method. Because digital images are pixel-wise discretized, binary images typically contain a large number of noisy pixels, requiring denoising to remove the noisy pixels and discontinuities along the boundary of the object. A series of morphological operations are performed on the binarized image, including image erosion, dilation, hole filling, etc. An unrecognizable object, such as a device or field environment, is also removed, typically in an area that appears closer to the image boundary.
(2) Three-dimensional volume reconstruction algorithm-three-dimensional volume reconstruction algorithm based on orthogonal calibration and volume correction
a) Least square method orthogonal calibration
Orthogonal calibration is required due to photogrammetry errors. The normalized orthonormal dimension is obtained from the following set of linear equations:
Ax＝b
the linear system solution minimizes the residual term, i.e
b) Volume correction
The volume of the segmented object may be calculated as a "voxel", i.e. a three-dimensional cuboid pixel. The volume of the reconstructed body can be obtained by voxel ratio between the rock and the calibration sphere, but the volume of the reconstructed object is always larger than or equal to the volume of the actual object. To correct for the larger volume after reconstruction, a correction factor was used, which was taken to be 0.95 after extensive experimental summary.
c) Resolution correction
Since the segmentation algorithm is based on the front-background contrast, the detected boundary will be slightly smaller than the actual object boundary. This result can lead to resolution-based overestimation, which is controlled by two parameters. The first parameter is the relative size ratio of the rock to the calibration sphere, and the second parameter is the absolute pixel occupancy of the calibration sphere. The following correction coefficient equation is used:
after a large number of experiments, the value of the correction parameter is generally 0.90.
And an output module 26, configured to output the rock geometric parameter feature.
Specifically, in the embodiment of the present invention, after the rock geometric parameter feature is calculated by the preset image segmentation algorithm, the result value expected by the embodiment of the present invention is obtained, and then the rock geometric parameter feature may be output to the display section or the storage end, so that the user may conveniently view and apply the rock geometric parameter feature.
Optionally, after the outputting the rock geometric parameter feature, the method further comprises: and integrating the preset segmentation algorithm to an application end with a user interface.
Specifically, in order to enable an application end with a user interface to obtain an accurate preset segmentation algorithm, the preset segmentation algorithm (algorithm for calculating geometric parameter characteristics of rock) displayed or output in the embodiment of the invention is integrated, that is, fused to a corresponding application end, so that large-size rock identification can be continued later.
According to another aspect of the embodiment of the present invention, there is also provided a nonvolatile storage medium, where the nonvolatile storage medium includes a stored program, and when the program runs, the program controls a device in which the nonvolatile storage medium is located to execute a large-size rock particle feature identification method.
Specifically, the method comprises the following steps: acquiring rock image data; generating a preset segmentation algorithm according to the rock image data; identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics; and outputting the geometric parameter characteristics of the rock.
According to another aspect of the embodiment of the present invention, there is also provided an electronic device including a processor and a memory; the memory stores computer readable instructions, and the processor is configured to execute the computer readable instructions, where the computer readable instructions execute a large-size rock particle feature recognition method when executed.
Specifically, the method comprises the following steps: acquiring rock image data; generating a preset segmentation algorithm according to the rock image data; identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics; and outputting the geometric parameter characteristics of the rock.
Through the embodiment, the technical problems that the manual measurement and the image analysis method in the prior art are not accurate and efficient for analyzing the large-size rock are solved.
Example III
As shown in fig. 4, for the identification of piled rock, i.e. a larger number of rock piles, the following technical matters may be performed to realize the third embodiment:
step S301, in-situ imaging: calibration spheres are typically used as standard reference objects to facilitate estimation of volume/size. The use of a calibration sphere has two options, 1) during image acquisition, if the camera/smartphone position is fixed, the calibration sphere can be captured first, then any object is captured, 2) when the camera/smartphone position is changing or the number of available devices is limited, the calibration sphere should be captured together with the object in each image to maintain consistency of reference. The first option is generally more efficient, but the latter is more widely used if the desired conditions are not met. Multiple views of the same aggregate were taken with a calibration racket and the cameras were placed in parallel on the slope of the aggregate. The camera requirement to acquire a sufficient resolution image is 2400 x 3000 or higher. Most smart phone cameras will pass this requirement. Aggregate images are taken from a direction nearly perpendicular to the aggregate surface, and aggregate images are taken from a location near the aggregate, so that most images are filled with useful rock pixels, while the calibration sphere (if present) does not appear too small in the image. To develop a robust algorithm, images with and without calibration balls are taken for training and segmentation purposes. For all images with calibration spheres, the sphere is located approximately in the center of the image, thus minimizing distortion effects. In addition, embodiments of the present invention select rocks with a particular color/texture because they contribute to the robustness of the data set.
Step S302, stack image segmentation: (1) The aggregate image segmentation algorithm comprises an image segmentation module based on deep learning and a morphological analysis module for particle shape characterization. The process of deep learning comprises three parts of preparation, training and analysis. (2) Constructing a rock stacking image marking dataset, wherein the dataset comprises aggregates from different areas and different lithologies, and particles in the aggregates comprise particles with different sizes, colors and textures; the data image includes photographs taken from different perspectives. For the sorted data set, the positions and the areas of all particles in each image are manually marked, and VGG Image Annotator (VIA) can simplify the manual marking work and further improve the working efficiency. These labeled regions are given a "Rock" label, and the neural network will search for this label and locate each particle region in the image.
Note that, the marking criteria in the above embodiment refer to: (1) The polygonal line should carefully approximate the particle boundaries with little deviation from the true shape (2) try to mark all identifiable particles except for very small particles that cannot be clearly identified by the naked eye and particles that cannot be resolved in dark areas; (3) The incomplete particles at the image boundary are marked, so that the segmentation model shows consistent performance at different positions in the image.
Specifically, the neural network architecture of the embodiment of the invention: the algorithm adopts a Mask R-CNN network, and the image segmentation task comprises a target detection step and a semantic segmentation step, so the network consists of two sub-networks, namely a regional convolutional neural network (R-CNN) for target detection and a full convolutional neural network (FCN) for semantic segmentation.
a) R-CNN network: the model first generates a large number of regions of interest (RoI) or region suggestions using a region suggestion network (RPN), then compresses each region suggestion into a feature map through a conventional CNN-based feature extraction network, the object classification model inputs the feature map into a linear Support Vector Machine (SVM), and reports the object classification and confidence of each region using non-maximal suppression. And merging the overlapped bounding boxes into a final bounding box at a position with higher confidence, and marking the final bounding box as a detection object.
b) FCN network: after the semantic segmentation target detection is carried out, the semantic segmentation is required to be carried out, and effective aggregation pixels in each boundary box are further extracted to obtain the shape and the boundary of the particle. The network is a network of a pure convolution layer and a pooling layer, and consists of a convolution network and a symmetrical deconvolution network. Through forward reasoning and backward propagation mechanisms, the trained network can take input images with any size and output local target areas of specified classes.
Step S3021, training for a neural network: the neural network is trained using the labeled aggregate image dataset. The process comprises two steps of forward transmission and backward propagation, wherein an input image is input into the neural network by the forward transmission, output is generated in the form of a segmented image, and in the backward propagation step, model parameters of the neural network are updated according to forward transmission errors, so that the neural network can realize self-adjustment or 'learning' to ensure that the segmented task is processed more accurately.
Step S303, particle morphology analysis algorithm: after successful segmentation, each region belonging to different polymeric particles is input into a morphological analysis module, and the equivalent size and slenderness ratio of each particle are calculated and expressed in the form of a histogram and cumulative distribution. The equivalent size of the particles used follows the definition of Equivalent Spherical Diameter (ESD), which is generally used to characterize the size of irregularly shaped objects as follows:
wherein A is the measurement area of the irregularly shaped object.
In the slenderness ratio (FER) calculation, the feret dimension is used to measure the shape of particles in a specified direction. The feret dimension, also known as caliper diameter, is defined as the distance between two parallel planes. Calculation of the slenderness ratio (FER) requires finding the maxima and minima of the feret dimension. Fei Te dimension, which is the largest or longest, is first obtained by searching the longest section of the particle area in all possible directionsDistance is determined. Then, by searching the edge is orthogonal to + ->The intercept of the direction gives the smallest or shortest Fei Te dimension +.>. FER is defined as the ratio between the maximum and minimum dimensions:
the foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In the several embodiments provided in the present application, it should be understood that the disclosed technology may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, for example, may be a logic function division, and may be implemented in another manner, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention.
Claims (1)
1. A flow chart of a large-size rock particle feature recognition method, which is characterized in that: the method comprises the following steps:
step S102, rock image data are acquired: for a single-particle rock sample, in order to analyze large-size rock data, corresponding large-size rock sample image data is firstly required to be acquired, and processing, analysis and treatment are carried out according to the acquired rock image data; after the rock image data are acquired, acquiring environment data, wherein the environment data comprise weather data and light data, and the environment data are acquired by acquiring and outputting the surrounding environment through image acquisition equipment or calling the surrounding environment through remote server data, so that the local weather environment data at the moment are determined; the rock image data are image data acquired through a plurality of image acquisition terminals;
firstly, acquiring corresponding large-size rock sample image data, and carrying out further processing, analysis and treatment according to the acquired rock image data; after the acquiring rock image data, the method further comprises: acquiring environment data: after the image data of the rock are acquired, the image recognition and analysis are more accurate by adding environmental data factors, and the environmental data are acquired by acquiring and outputting the surrounding environment through image acquisition equipment or determining the local environmental data at the moment through remote server data call;
The rock image data is image data acquired by a plurality of image acquisition terminals: the image acquisition apparatus includes: three smart phones, a high resolution camera and a remote shutter control, three copper tubes with the length of 1.52 meters, which are spliced by a short pipeline by connectors, ensure that the copper tubes are easy to assemble and disassemble, a 10 kg patio umbrella base is used as an anchorage of the copper tubes, three blue curtains with the length of 1.52 meters multiplied by 1.52 meters are used as a background and a bottom surface, and three camera tripods are used for fixing the smart phones on the top/front/side surfaces; one tripod is provided with a cantilever; after equipment debugging is completed, rock photos with different sizes are shot from different observation angles to form a database; after each shooting is completed, randomly rotating the particles by an angle to enable each particle to be repeatedly shot at least three times; the accuracy of the size and shape measurements was verified by comparison with manually measuring the size and weight of individual polished rocks;
step S104, a preset segmentation algorithm is generated according to the rock image data; the preset segmentation algorithm is an image segmentation algorithm based on multi-view information, and the reconstruction quantization is carried out on the volume of the object by combining a three-dimensional reconstruction algorithm;
The image segmentation algorithm based on the multi-view information comprises the following steps:
(1) The image segmentation algorithm is a color-based target detection image segmentation algorithm, and comprises three parts, namely color space representation, foreground and background contrast enhancement, self-adaptive threshold value and morphological denoising;
a) Color space representation—cie L a b color space: in the L x a x b x space of the CIE, L x channels represent luminance or intensity values, a x channels and b x channels track the green to red, blue to yellow transitions, respectively; the CIE L x a x b x space eliminates shadowing well by separating useful information into its a x and b x channels; depending on the object color and background color, useful object information is accumulated in the a-channel, b-channel, or both;
b) Foreground-background contrast enhancement: firstly, obtaining pixel histograms of an a channel and a b channel, and obtaining a pixel cumulative distribution function through the pixel histograms;
the contrast enhancement of the foreground and the background is realized through the color distance, the gray level intensity of the pixel represents the distance of the color taking the foreground representing color as a reference, and the closer the pixel color is to the foreground representing color, the smaller the color distance in the distance graph is, the darker the intensity is, and vice versa;
c) Adaptive thresholding and morphological denoising
Based on the enhanced distance map, image thresholding is applied to obtain a binarized image, and the threshold is set as a fixed threshold or an adaptive threshold; because the digital image is discretized by pixels, the binary image contains a large number of noise pixels, and denoising is needed to remove the noise pixels and discontinuous points along the boundary of the target; carrying out morphological operation on the binarized image, wherein the morphological operation comprises image erosion, expansion and hole filling; deleting unrecognizable objects closer to the image boundary;
(2) Three-dimensional volume reconstruction algorithm-three-dimensional volume reconstruction algorithm based on orthogonal calibration and volume correction
a) Least square method orthogonal calibration
Because the photogrammetry error requires orthogonal calibration, a normalized orthogonal correction dimension is obtained according to the following linear equation set:
Ax＝b
；
its solution minimizes the residual term, i.e
；
b) Volume correction
The volume of the segmented object is calculated with "voxels", i.e. three-dimensional cuboid pixels; obtaining a reconstruction body from the voxel ratio between the rock and the calibration sphere, wherein the volume of the reconstructed object is always larger than or equal to that of the actual object; to correct the reconstructed larger volume, a correction coefficient is adopted, wherein the correction coefficient is 0.95;
c) Resolution correction
Since the segmentation algorithm is based on pre-background contrast, the detected boundary will be slightly smaller than the actual object boundary, which results in resolution-based overestimation, which is governed by two parameters; the first parameter is the relative size ratio of the rock to the calibration sphere, and the second parameter is the absolute pixel occupancy of the calibration sphere; the following correction coefficient equation is used:
；
the value of the correction parameter is 0.90;
step S106, identifying the rock image data according to the preset segmentation algorithm to obtain rock geometric parameter characteristics;
step S108, outputting the rock geometric parameter characteristics;
after the rock geometric parameter characteristics are calculated through a preset image segmentation algorithm, the obtained result value is output to a display section or a storage end, so that a user can conveniently check and apply the rock geometric parameter characteristics; after outputting the rock geometry parameter feature, the method further comprises: integrating the preset segmentation algorithm to an application end with a user interface; in order to enable the application end with the user interface to obtain an accurate preset segmentation algorithm, integrating the displayed or output preset segmentation algorithm, namely fusing the preset segmentation algorithm to the corresponding application end so as to facilitate subsequent continuous large-size rock identification;
For stacked rock, the method further comprises:
step S301, in-situ imaging: the calibration sphere is used as a standard reference object to facilitate estimating the volume/size of the rock, and there are two options for the use of the calibration sphere 1) during image acquisition, if the camera/smartphone position is fixed, the calibration sphere is captured first, then the rock object is captured; 2) When the position of the camera/smartphone is constantly changing or the number of available devices is limited, the calibration sphere should be captured together with the object in each image to maintain consistency of the reference; shooting multiple views of the same aggregate with a calibration racket, and placing cameras in parallel on the slope of the aggregate; the camera requirement to acquire a sufficient resolution image is 2400x3000; aggregate images are taken from a direction nearly perpendicular to the aggregate surface, the aggregate images being taken from a location near the aggregate;
step S302, stack image segmentation: (1) The aggregate image segmentation algorithm comprises an image segmentation module based on deep learning and a morphology analysis module for particle shape characterization based on the flow of the deep learning; the process of deep learning comprises three parts of preparation, training and analysis; (2) Constructing a rock stacking image marking dataset, wherein the dataset comprises aggregates from different areas and different lithologies, and particles in the aggregates comprise particles with different sizes, colors and textures; the data image includes photographs taken from different perspectives; for the collated dataset, the positions and areas of all particles in each image were manually marked, using VGG Image Annotator (VIA) to simplify the manual marking effort; these labeled regions are given a "Rock" label, which the neural network will search for and locate each particle region in the image;
The criteria for the marking include: (1) The polygonal line should approximate the particle boundary and have little deviation from the true shape; (2) trying to mark all identifiable particles; (3) Marking incomplete particles at the boundary of the image to enable the segmentation model to show consistent performance at different positions in the image;
architecture of neural network: the aggregate image segmentation algorithm adopts a Mask R-CNN network, and the image segmentation task comprises a target detection step and a semantic segmentation step, so that the network consists of two sub-networks, namely a regional convolution neural network (R-CNN) for target detection and a full convolution neural network (FCN) for semantic segmentation;
a) R-CNN network: the R-CNN network firstly generates a large number of regions of interest (RoI) or region suggestions by using a region suggestion network (RPN), then compresses each region suggestion into a feature map by using a traditional CNN-based feature extraction network, inputs the feature map into a linear Support Vector Machine (SVM) by using a target classification model, and reports the target classification and the confidence of each region by using non-maximum suppression; merging the overlapped bounding boxes into a final bounding box at a position with higher confidence, and marking the final bounding box as a detection object;
b) FCN network: after the FCN network performs semantic segmentation target detection, semantic segmentation is required to be performed, and effective aggregation pixels in each boundary box are further extracted to obtain the shape and the boundary of the particle; the FCN network is a network of a pure convolution layer and a pooling layer, and consists of a convolution network and a symmetrical deconvolution network; the trained network takes an input image with any size through forward reasoning and backward propagation mechanisms, and outputs a local target area of a specified class;
step S3021, training for a neural network: training the Mask R-CNN network using the marked aggregate image dataset; the process comprises two steps of forward transmission and backward transmission, wherein an input image is input into a neural network by the forward transmission, output is generated in a segmented image form, and in the backward transmission step, model parameters of the neural network are updated according to forward transmission errors, so that the neural network can realize self-adjustment or 'learning' to ensure that segmentation tasks are processed more accurately;
step S303, particle morphology analysis algorithm: after successful segmentation, inputting each region belonging to different aggregation particles into a morphological analysis module, calculating the equivalent size and slenderness ratio of each particle, and representing the equivalent size and slenderness ratio in the form of a histogram and cumulative distribution; the equivalent size of the particles used follows the definition of Equivalent Spherical Diameter (ESD), which is used to characterize the size of irregularly shaped objects as follows:
；
Wherein A is the measurement area of an irregularly shaped object;
in the calculation of the slenderness ratio, the shape of the particle in a given direction is measured by the Ferrett dimension, also known as the caliper, defined as the distance between two parallel planes, the calculation of the slenderness ratio (FER) requires finding the maxima and minima of the Ferrett dimension, L max The largest or longest Fei Te dimension is determined by first searching the longest intercept of the particle region in all possible directions; then, by searching along the orthogonal L max The intercept of the direction results in a minimum or shortest Fei Te dimension L min The method comprises the steps of carrying out a first treatment on the surface of the FER is defined as the ratio between the maximum and minimum dimensions:
。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110410018.9A CN113177949B (en) | 2021-04-16 | 2021-04-16 | Large-size rock particle feature recognition method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110410018.9A CN113177949B (en) | 2021-04-16 | 2021-04-16 | Large-size rock particle feature recognition method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113177949A CN113177949A (en) | 2021-07-27 |
| CN113177949B true CN113177949B (en) | 2023-09-01 |
Family
ID=76923823
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110410018.9A Active CN113177949B (en) | 2021-04-16 | 2021-04-16 | Large-size rock particle feature recognition method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113177949B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113624163B (en) * | 2021-08-11 | 2022-04-15 | 西南交通大学 | Three-dimensional laser scanning-based gravel particle surface edge angle measurement method |
Citations (54)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556703A (en) * | 2009-05-16 | 2009-10-14 | 中国石油大学(华东) | Method for establishing network model based on serial section image |
| CN102074012A (en) * | 2011-01-22 | 2011-05-25 | 四川农业大学 | Method for three-dimensionally reconstructing tender shoot state of tea by combining image and computation model |
| CN102867302A (en) * | 2012-08-30 | 2013-01-09 | 四川大学 | Core fracture identification method based on three-dimensional image information processing |
| CN103047942A (en) * | 2012-12-26 | 2013-04-17 | 浙江大学 | Visual acquisition system and method for geometrical characteristics of graded crushed rocks of railway and road beds |
| CN103063134A (en) * | 2012-12-26 | 2013-04-24 | 浙江大学 | Gravel geometric feature acquisition system and acquisition method |
| CN203100694U (en) * | 2012-12-26 | 2013-07-31 | 浙江大学 | Visual acquisition system for geometrical characteristics of railway and highway bed graded broken stone |
| CN203100677U (en) * | 2012-12-26 | 2013-07-31 | 浙江大学 | Gravel geometric feature acquisition system |
| CN103325117A (en) * | 2013-06-17 | 2013-09-25 | 中国石油天然气集团公司 | Rock core picture processing method and system based on MATLAB |
| CN103325118A (en) * | 2013-06-26 | 2013-09-25 | 中国石油大学(北京) | Method and device for acquiring characteristic parameters of core hole of carbonatite |
| CN103759645A (en) * | 2014-01-26 | 2014-04-30 | 湖南航天机电设备与特种材料研究所 | Aerostat capsule volume measurement method |
| CN103942803A (en) * | 2014-05-05 | 2014-07-23 | 北京理工大学 | SAR (Synthetic Aperture Radar) image based automatic water area detection method |
| CN103971410A (en) * | 2014-05-23 | 2014-08-06 | 中国石油大学(华东) | Three-dimensional rock core visualization method based on CT images |
| CN104182759A (en) * | 2014-08-20 | 2014-12-03 | 徐州坤泰电子科技有限公司 | Scanning electron microscope based particle morphology identification method |
| CN104502245A (en) * | 2015-01-08 | 2015-04-08 | 青岛理工大学 | Method for determining fineness modulus of sand by utilizing image analysis technology |
| CN205451195U (en) * | 2016-03-07 | 2016-08-10 | 深圳市鹰眼在线电子科技有限公司 | Real -time three -dimensional some cloud system that rebuilds based on many cameras |
| CN106548480A (en) * | 2016-12-23 | 2017-03-29 | 蚌埠学院 | A kind of agricultural product volume rapid measurement device and measuring method based on machine vision |
| CN106651825A (en) * | 2015-11-03 | 2017-05-10 | 中国科学院沈阳计算技术研究所有限公司 | Workpiece positioning and identification method based on image segmentation |
| CN106682597A (en) * | 2016-12-14 | 2017-05-17 | 四川长虹电器股份有限公司 | Method for identifying traffic lights |
| CN106845453A (en) * | 2017-02-24 | 2017-06-13 | 太原理工大学 | Taillight detection and recognition methods based on image |
| CN206331487U (en) * | 2016-12-23 | 2017-07-14 | 蚌埠学院 | A kind of agricultural product volume rapid measurement device based on machine vision |
| CN107507145A (en) * | 2017-08-25 | 2017-12-22 | 上海海洋大学 | A kind of underwater picture Enhancement Method based on the stretching of different colours spatially adaptive histogram |
| CN107655796A (en) * | 2017-09-18 | 2018-02-02 | 青岛理工大学 | A kind of building sand modulus of fineness rapid assay methods |
| CN107891012A (en) * | 2017-10-11 | 2018-04-10 | 杭州电子科技大学 | Pearl size and circularity sorting equipment based on equivalent algorithm |
| CN108305263A (en) * | 2018-01-31 | 2018-07-20 | 中国科学院武汉岩土力学研究所 | A kind of image statistics method of rock fragmentation after explosion |
| CN108550154A (en) * | 2018-04-11 | 2018-09-18 | 中国科学院西双版纳热带植物园 | A kind of method of accurately measuring karst earth's surface bare rock accounting |
| CN108596853A (en) * | 2018-04-28 | 2018-09-28 | 上海海洋大学 | Underwater picture Enhancement Method based on bias light statistical model and transmission map optimization |
| CN108645717A (en) * | 2018-03-09 | 2018-10-12 | 绍兴文理学院 | A kind of petrophysical model structural plane shearing test overall process method for visualizing |
| CN108765443A (en) * | 2018-05-22 | 2018-11-06 | 杭州电子科技大学 | A kind of mark enhancing processing method of adaptive color Threshold segmentation |
| CN108830877A (en) * | 2018-06-08 | 2018-11-16 | 中国石油化工股份有限公司 | The stereoscopic micro-image quantitative description of rock core |
| CN109255775A (en) * | 2018-07-18 | 2019-01-22 | 苏州精观医疗科技有限公司 | A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically |
| CN109544624A (en) * | 2018-11-09 | 2019-03-29 | 北方爆破科技有限公司 | A kind of rock fragmentation image analysis system |
| CN109544467A (en) * | 2018-10-23 | 2019-03-29 | 江苏理工学院 | A method of based on enhancing color image contrast under LAB model |
| CN109948712A (en) * | 2019-03-20 | 2019-06-28 | 天津工业大学 | A kind of nanoparticle size measurement method based on improved Mask R-CNN |
| CN110018108A (en) * | 2019-05-14 | 2019-07-16 | 中国科学院地质与地球物理研究所 | The measuring method and equipment in blowhole aperture |
| CN110033516A (en) * | 2019-04-16 | 2019-07-19 | 重庆交通大学 | Flat-elongated particles detection method of content based on the identification of binocular camera Image Acquisition |
| CN110095388A (en) * | 2019-04-18 | 2019-08-06 | 中国石油大学(北京) | The determination method and device of clastic rock grain structure |
| CN110119753A (en) * | 2019-01-08 | 2019-08-13 | 长江岩土工程总公司（武汉） | A kind of method of reconstituted texture identification lithology |
| CN110230978A (en) * | 2018-03-14 | 2019-09-13 | 浙江大学山东工业技术研究院 | A kind of refractory brick geometric dimension measurement method |
| CN110570401A (en) * | 2019-08-19 | 2019-12-13 | 成都理工大学 | Slope fractured rock mass information acquisition method based on image recognition |
| CN110660060A (en) * | 2019-11-29 | 2020-01-07 | 武汉精立电子技术有限公司 | OLED screen color mixing defect detection method, terminal equipment and computer readable medium |
| CN110751620A (en) * | 2019-08-28 | 2020-02-04 | 宁波海上鲜信息技术有限公司 | Method for estimating volume and weight, electronic device, and computer-readable storage medium |
| CN110849324A (en) * | 2019-10-29 | 2020-02-28 | 中铁十二局集团有限公司 | Long exposure oblique photography tunnel holographic measurement method |
| CN110907317A (en) * | 2018-09-14 | 2020-03-24 | 中石化石油工程技术服务有限公司 | Rock particle analysis method, device, equipment and computer readable storage equipment |
| CN111028181A (en) * | 2019-12-25 | 2020-04-17 | 歌尔股份有限公司 | Image enhancement processing method, device, equipment and storage medium |
| CN111079773A (en) * | 2019-12-27 | 2020-04-28 | 中国地质调查局成都地质调查中心 | Gravel parameter acquisition method, device, equipment and storage medium based on Mask R-CNN network |
| CN111161292A (en) * | 2019-11-21 | 2020-05-15 | 合肥合工安驰智能科技有限公司 | Ore size measurement method and application system |
| CN111380785A (en) * | 2020-03-30 | 2020-07-07 | 中南大学 | Two-dimensional geometric characteristic parameter acquisition system and method for rock particles |
| CN111398273A (en) * | 2019-10-09 | 2020-07-10 | 天津大学 | Rock geometric-mechanical parameter acquisition method and holographic scanning system |
| CN111524229A (en) * | 2020-03-30 | 2020-08-11 | 中南大学 | Three-dimensional geometric morphology information extraction system and method for rock particles |
| CN112001878A (en) * | 2020-05-21 | 2020-11-27 | 合肥合工安驰智能科技有限公司 | Deep learning ore scale measuring method based on binarization neural network and application system |
| CN112424829A (en) * | 2018-06-29 | 2021-02-26 | 沙特阿拉伯石油公司 | Identification of geometrical properties of rock structures by digital imaging |
| CN112419250A (en) * | 2020-11-13 | 2021-02-26 | 湖北工业大学 | Pavement crack digital image extraction, crack repair and crack parameter calculation method |
| CN112528904A (en) * | 2020-12-18 | 2021-03-19 | 中山艾尚智同信息科技有限公司 | Image segmentation method for sand particle size detection system |
| CN112529948A (en) * | 2020-12-25 | 2021-03-19 | 南京林业大学 | Mature pomegranate positioning method based on Mask R-CNN and 3-dimensional sphere fitting |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3033068B1 (en) * | 2015-02-24 | 2017-03-10 | Ifp Energies Now | METHOD OF SEGMENTING THE IMAGE OF A RECONSTRUCTED OBJECT BY THREE DIMENSIONAL RECONSTRUCTION |
| US20190138786A1 (en) * | 2017-06-06 | 2019-05-09 | Sightline Innovation Inc. | System and method for identification and classification of objects |
-
2021
- 2021-04-16 CN CN202110410018.9A patent/CN113177949B/en active Active
Patent Citations (54)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556703A (en) * | 2009-05-16 | 2009-10-14 | 中国石油大学(华东) | Method for establishing network model based on serial section image |
| CN102074012A (en) * | 2011-01-22 | 2011-05-25 | 四川农业大学 | Method for three-dimensionally reconstructing tender shoot state of tea by combining image and computation model |
| CN102867302A (en) * | 2012-08-30 | 2013-01-09 | 四川大学 | Core fracture identification method based on three-dimensional image information processing |
| CN103047942A (en) * | 2012-12-26 | 2013-04-17 | 浙江大学 | Visual acquisition system and method for geometrical characteristics of graded crushed rocks of railway and road beds |
| CN103063134A (en) * | 2012-12-26 | 2013-04-24 | 浙江大学 | Gravel geometric feature acquisition system and acquisition method |
| CN203100694U (en) * | 2012-12-26 | 2013-07-31 | 浙江大学 | Visual acquisition system for geometrical characteristics of railway and highway bed graded broken stone |
| CN203100677U (en) * | 2012-12-26 | 2013-07-31 | 浙江大学 | Gravel geometric feature acquisition system |
| CN103325117A (en) * | 2013-06-17 | 2013-09-25 | 中国石油天然气集团公司 | Rock core picture processing method and system based on MATLAB |
| CN103325118A (en) * | 2013-06-26 | 2013-09-25 | 中国石油大学(北京) | Method and device for acquiring characteristic parameters of core hole of carbonatite |
| CN103759645A (en) * | 2014-01-26 | 2014-04-30 | 湖南航天机电设备与特种材料研究所 | Aerostat capsule volume measurement method |
| CN103942803A (en) * | 2014-05-05 | 2014-07-23 | 北京理工大学 | SAR (Synthetic Aperture Radar) image based automatic water area detection method |
| CN103971410A (en) * | 2014-05-23 | 2014-08-06 | 中国石油大学(华东) | Three-dimensional rock core visualization method based on CT images |
| CN104182759A (en) * | 2014-08-20 | 2014-12-03 | 徐州坤泰电子科技有限公司 | Scanning electron microscope based particle morphology identification method |
| CN104502245A (en) * | 2015-01-08 | 2015-04-08 | 青岛理工大学 | Method for determining fineness modulus of sand by utilizing image analysis technology |
| CN106651825A (en) * | 2015-11-03 | 2017-05-10 | 中国科学院沈阳计算技术研究所有限公司 | Workpiece positioning and identification method based on image segmentation |
| CN205451195U (en) * | 2016-03-07 | 2016-08-10 | 深圳市鹰眼在线电子科技有限公司 | Real -time three -dimensional some cloud system that rebuilds based on many cameras |
| CN106682597A (en) * | 2016-12-14 | 2017-05-17 | 四川长虹电器股份有限公司 | Method for identifying traffic lights |
| CN206331487U (en) * | 2016-12-23 | 2017-07-14 | 蚌埠学院 | A kind of agricultural product volume rapid measurement device based on machine vision |
| CN106548480A (en) * | 2016-12-23 | 2017-03-29 | 蚌埠学院 | A kind of agricultural product volume rapid measurement device and measuring method based on machine vision |
| CN106845453A (en) * | 2017-02-24 | 2017-06-13 | 太原理工大学 | Taillight detection and recognition methods based on image |
| CN107507145A (en) * | 2017-08-25 | 2017-12-22 | 上海海洋大学 | A kind of underwater picture Enhancement Method based on the stretching of different colours spatially adaptive histogram |
| CN107655796A (en) * | 2017-09-18 | 2018-02-02 | 青岛理工大学 | A kind of building sand modulus of fineness rapid assay methods |
| CN107891012A (en) * | 2017-10-11 | 2018-04-10 | 杭州电子科技大学 | Pearl size and circularity sorting equipment based on equivalent algorithm |
| CN108305263A (en) * | 2018-01-31 | 2018-07-20 | 中国科学院武汉岩土力学研究所 | A kind of image statistics method of rock fragmentation after explosion |
| CN108645717A (en) * | 2018-03-09 | 2018-10-12 | 绍兴文理学院 | A kind of petrophysical model structural plane shearing test overall process method for visualizing |
| CN110230978A (en) * | 2018-03-14 | 2019-09-13 | 浙江大学山东工业技术研究院 | A kind of refractory brick geometric dimension measurement method |
| CN108550154A (en) * | 2018-04-11 | 2018-09-18 | 中国科学院西双版纳热带植物园 | A kind of method of accurately measuring karst earth's surface bare rock accounting |
| CN108596853A (en) * | 2018-04-28 | 2018-09-28 | 上海海洋大学 | Underwater picture Enhancement Method based on bias light statistical model and transmission map optimization |
| CN108765443A (en) * | 2018-05-22 | 2018-11-06 | 杭州电子科技大学 | A kind of mark enhancing processing method of adaptive color Threshold segmentation |
| CN108830877A (en) * | 2018-06-08 | 2018-11-16 | 中国石油化工股份有限公司 | The stereoscopic micro-image quantitative description of rock core |
| CN112424829A (en) * | 2018-06-29 | 2021-02-26 | 沙特阿拉伯石油公司 | Identification of geometrical properties of rock structures by digital imaging |
| CN109255775A (en) * | 2018-07-18 | 2019-01-22 | 苏州精观医疗科技有限公司 | A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically |
| CN110907317A (en) * | 2018-09-14 | 2020-03-24 | 中石化石油工程技术服务有限公司 | Rock particle analysis method, device, equipment and computer readable storage equipment |
| CN109544467A (en) * | 2018-10-23 | 2019-03-29 | 江苏理工学院 | A method of based on enhancing color image contrast under LAB model |
| CN109544624A (en) * | 2018-11-09 | 2019-03-29 | 北方爆破科技有限公司 | A kind of rock fragmentation image analysis system |
| CN110119753A (en) * | 2019-01-08 | 2019-08-13 | 长江岩土工程总公司（武汉） | A kind of method of reconstituted texture identification lithology |
| CN109948712A (en) * | 2019-03-20 | 2019-06-28 | 天津工业大学 | A kind of nanoparticle size measurement method based on improved Mask R-CNN |
| CN110033516A (en) * | 2019-04-16 | 2019-07-19 | 重庆交通大学 | Flat-elongated particles detection method of content based on the identification of binocular camera Image Acquisition |
| CN110095388A (en) * | 2019-04-18 | 2019-08-06 | 中国石油大学(北京) | The determination method and device of clastic rock grain structure |
| CN110018108A (en) * | 2019-05-14 | 2019-07-16 | 中国科学院地质与地球物理研究所 | The measuring method and equipment in blowhole aperture |
| CN110570401A (en) * | 2019-08-19 | 2019-12-13 | 成都理工大学 | Slope fractured rock mass information acquisition method based on image recognition |
| CN110751620A (en) * | 2019-08-28 | 2020-02-04 | 宁波海上鲜信息技术有限公司 | Method for estimating volume and weight, electronic device, and computer-readable storage medium |
| CN111398273A (en) * | 2019-10-09 | 2020-07-10 | 天津大学 | Rock geometric-mechanical parameter acquisition method and holographic scanning system |
| CN110849324A (en) * | 2019-10-29 | 2020-02-28 | 中铁十二局集团有限公司 | Long exposure oblique photography tunnel holographic measurement method |
| CN111161292A (en) * | 2019-11-21 | 2020-05-15 | 合肥合工安驰智能科技有限公司 | Ore size measurement method and application system |
| CN110660060A (en) * | 2019-11-29 | 2020-01-07 | 武汉精立电子技术有限公司 | OLED screen color mixing defect detection method, terminal equipment and computer readable medium |
| CN111028181A (en) * | 2019-12-25 | 2020-04-17 | 歌尔股份有限公司 | Image enhancement processing method, device, equipment and storage medium |
| CN111079773A (en) * | 2019-12-27 | 2020-04-28 | 中国地质调查局成都地质调查中心 | Gravel parameter acquisition method, device, equipment and storage medium based on Mask R-CNN network |
| CN111380785A (en) * | 2020-03-30 | 2020-07-07 | 中南大学 | Two-dimensional geometric characteristic parameter acquisition system and method for rock particles |
| CN111524229A (en) * | 2020-03-30 | 2020-08-11 | 中南大学 | Three-dimensional geometric morphology information extraction system and method for rock particles |
| CN112001878A (en) * | 2020-05-21 | 2020-11-27 | 合肥合工安驰智能科技有限公司 | Deep learning ore scale measuring method based on binarization neural network and application system |
| CN112419250A (en) * | 2020-11-13 | 2021-02-26 | 湖北工业大学 | Pavement crack digital image extraction, crack repair and crack parameter calculation method |
| CN112528904A (en) * | 2020-12-18 | 2021-03-19 | 中山艾尚智同信息科技有限公司 | Image segmentation method for sand particle size detection system |
| CN112529948A (en) * | 2020-12-25 | 2021-03-19 | 南京林业大学 | Mature pomegranate positioning method based on Mask R-CNN and 3-dimensional sphere fitting |
Non-Patent Citations (1)
| Title |
|---|
| 阳波 ; 段吉安 ; 郑煜 ; .石料加工过程中颗粒粒性表示与计算.计算机工程与应用.2009,(第10期),第202-203+207页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113177949A (en) | 2021-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kersten et al. | Image-based low-cost systems for automatic 3D recording and modelling of archaeological finds and objects | |
| Graham et al. | Automated sizing of coarse-grained sediments: image-processing procedures | |
| Zhou et al. | Image-based 3D reconstruction for posthurricane residential building damage assessment | |
| US8331654B2 (en) | Stereo-image registration and change detection system and method | |
| CN106934795B (en) | A kind of automatic testing method and prediction technique of glue into concrete beam cracks | |
| Woodget et al. | From manned to unmanned aircraft: Adapting airborne particle size mapping methodologies to the characteristics of sUAS and SfM | |
| Moaveni et al. | Evaluation of aggregate size and shape by means of segmentation techniques and aggregate image processing algorithms | |
| Verma et al. | A method based on structure-from-motion photogrammetry to generate sub-millimetre-resolution digital elevation models for investigating rock breakdown features | |
| Daghigh et al. | A critical review of discontinuity plane extraction from 3D point cloud data of rock mass surfaces | |
| Dan et al. | Application of deep learning-based image recognition technology to asphalt–aggregate mixtures: Methodology | |
| Lee et al. | Semi-automatic calculation of joint trace length from digital images based on deep learning and data structuring techniques | |
| An et al. | A fast and practical method for determining particle size and shape by using smartphone photogrammetry | |
| Chehata et al. | LIDAR data classification using hierarchical K-means clustering | |
| CN113177949B (en) | Large-size rock particle feature recognition method and device | |
| Bartels et al. | Segmentation of LIDAR data using measures of distribution | |
| Tuan et al. | Automatic stereo vision-based inspection system for particle shape analysis of coarse aggregates | |
| Li et al. | 3D map system for tree monitoring in hong kong using google street view imagery and deep learning | |
| Markiewicz et al. | Quality assessment of the use of a medium format camera in the investigation of wall paintings: An image-based approach | |
| Anusree et al. | Characterization of sand particle morphology: state-of-the-art | |
| Ebrahimikia et al. | True orthophoto generation based on unmanned aerial vehicle images using reconstructed edge points | |
| Alshawabkeh et al. | A new true ortho‐photo methodology for complex archaeological application | |
| Zheng et al. | Accuracy comparison of rock discontinuity geometric parameters in photogrammetry based on two georeferencing methods: Control points and geotagged photos | |
| CN108335321B (en) | Automatic ground surface gravel size information extraction method based on multi-angle photos | |
| Seong et al. | Image-based 3D building reconstruction using A-KAZE feature extraction algorithm | |
| Budianti et al. | Background blurring and removal for 3d modelling of cultural heritage objects |
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 |