CN115753612A - Intelligent core box, core image acquisition system and intelligent core recording method - Google Patents
Intelligent core box, core image acquisition system and intelligent core recording method Download PDFInfo
- Publication number
- CN115753612A CN115753612A CN202211474459.6A CN202211474459A CN115753612A CN 115753612 A CN115753612 A CN 115753612A CN 202211474459 A CN202211474459 A CN 202211474459A CN 115753612 A CN115753612 A CN 115753612A
- Authority
- CN
- China
- Prior art keywords
- core
- image
- model
- box
- image acquisition
- 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.)
- Pending
Links
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention belongs to the technical field of engineering geological investigation, and particularly relates to an intelligent core box, a core image acquisition system and an intelligent core compiling and recording method. The intelligent core cataloguing method mainly comprises the following steps: s1, assembling a core image acquisition system; s2, correcting the camera; s3, collecting images of the core box or the core box group; s4, preprocessing an image; s5, establishing an image recognition model; s6, after the training of the drilling core image recognition model is finished in S5, selecting the picture processed in S4 as the drilling core image to be recognized, and guiding the picture into the recognition model for recognition; and S7, obtaining the core type and core depth information of each section of core column. The intelligent core cataloging method has the advantages that the lithology recognition and the drilling histogram mapping can be automatically and quickly completed, the device is simple to disassemble and assemble, the indoor operation time of reconnaissance personnel is saved, and the intelligent core cataloging method is very suitable for the field drilling core cataloging work.
Description
Technical Field
The invention belongs to the technical field of engineering geological investigation, and particularly relates to an intelligent core box, a core image acquisition system and an intelligent core compiling and recording method.
Background
The drilling exploration is one of important means for exploring the quality and the attribute of the engineering rock mass, and is widely used for engineering project exploration. In railway or traffic investigation projects, information such as lithology, structural surface geometry and the like can be obtained through drilling investigation, and the method has great significance for geological description and engineering design. The field reconnaissance personnel mostly rely on experience in the field to directly observe characteristics such as rock color, surface texture and hardness so as to identify the approximate name of the rock. The human eye observation method is simple and rapid, but is limited by the geological professional knowledge of the main body, and subjective judgment errors exist. The artificial intelligence algorithm provides an important way for lithology identification, and the efficiency and the accuracy of lithology identification are greatly improved. CN202110565704.3 discloses a drilling image intelligent identification method and device based on a neural network. However, this method can only photograph and identify a local area of the core, and cannot automatically scan and identify the core of a complete borehole, which makes the field application of the method poor.
The core box in the prior art is of a wood structure, is disposable and can only be used for storing cores. The existing invention patent only considers the convenience of storage and transportation of the rock core and can not complete the whole-hole rock core identification and cataloguing work.
Disclosure of Invention
The invention provides an intelligent core box, a core image acquisition system and an intelligent core editing and recording method.
The technical scheme provided by the invention has the following beneficial effects: the utility model provides a core box, includes the box body, the box body is the cuboid structure, and its upper end is uncovered and is set up, is equipped with the thing groove of putting of rectangle structure in it, one side upper end of box body is equipped with a plurality of interval distribution draw-in grooves, and the upper end of its relative one side is equipped with a plurality of the pothook of draw-in groove looks adaptation, be equipped with the mounting hole that is used for the installing support on the box body.
A core image acquisition system, includes support, image acquisition unit and at least one core box as in claim 1, the support is the mechanism of "L" shape of falling, and its vertical section detachably installs in the mounting hole, its horizontal segment extends to put the top of thing groove, the horizontal segment of support is equipped with the fixation clamp, the fixation clamp is used for fixed image acquisition unit, image acquisition unit is used for gathering the core image.
Further, the image acquisition unit is a mobile phone.
An intelligent core logging method is executed by the core image acquisition system and comprises the following steps:
s1, assembling a core image acquisition system, and placing a drill core into a core box;
s2, camera correction: setting mark points at four right angles of a box body of each core box, adjusting the focal length of the debugging image acquisition unit to enable the whole visual field to cover the core box, and simultaneously ensuring that the acquired image information of the core box is clear;
s3, after the core image acquisition system is assembled, shooting a core box by using the image acquisition unit, and acquiring an image of the core box, wherein if the core image acquisition system comprises a plurality of core boxes, an image of a core box group consisting of the plurality of core boxes is acquired;
s4, image preprocessing: cutting and segmenting the acquired core box image or core box group image to obtain a columnar image of a single drill hole core, and preprocessing the columnar image of the single drill hole core to obtain a preprocessed picture with a unified specification;
s5, establishing an image recognition model: establishing a drilling core image recognition model by using a deep learning algorithm based on a convolutional neural network and by using an Inception V3 pre-training model method, and performing model training by using the drilling core image subjected to image preprocessing in S4;
s6, after the drill hole core image recognition model is trained in the S5, selecting the image processed in the S4 as a drill hole core image to be recognized, and guiding the image into the recognition model for recognition;
and S7, after the image recognition is finished in the S6, respectively obtaining the core type and the core depth information of each section of core column according to the recognition result, marking the core depth corresponding to the core type, and establishing a core column CAD graph.
Further, in S1, if the core image acquisition system includes a plurality of core boxes, assembling and fixing the plurality of core boxes to obtain a core box group, fixing the bracket in the mounting hole of the box body located at the central position, and mounting the image acquisition unit on the fixing clamp; and debugging the image acquisition distance of the image acquisition unit, and placing a target rock core in each box body.
Further, the image preprocessing in S4 specifically includes:
s4.1, cutting the core box image or the core box group image by using an image segmentation technology, segmenting the core box image or the core box group image into a histogram of a single drilling core, and splicing the obtained histograms of a plurality of drilling cores if the number of core boxes is multiple;
s4.2, randomly intercepting the single or spliced drill core histogram according to the size of 320 multiplied by 320 pixels;
s4.3, turning and rotating the randomly captured picture;
s4.4, fine adjustment of brightness is carried out on the drill hole core image obtained through processing;
and S4.5, uniformly scaling all the obtained pictures to 256 multiplied by 256 pixels to obtain the preprocessed borehole core image.
Further, the specific steps of the image recognition model in S5 are:
s5.1, data preparation: collecting known types of drilling core images, processing the known types of drilling core images according to the method in the step 4, labeling each drilling core image, and establishing a drilling core model training set;
s5.2, model construction: completely extracting the convolutional layer and the pooling layer of the convolutional neural network Inception V3, removing the last output layer of the Inception V3 model, adding two initialized full-connected layers, and extracting the image characteristics of the drill core by adopting the parameter values of the Inception V3 model;
s5.3, model training: training the drill hole core recognition model by adopting the training sample obtained in the step S4;
s5.4, model testing: pushing the known borehole core images except the training set to preprocessing according to the method in the step 4, establishing a borehole core test set, and checking the identification accuracy of the model by using the test set; and if the model can be correctly identified, the model is qualified, and if the model cannot be correctly identified, the training sample is collected again to train the model and adjust the parameters of the model until the model can be correctly identified.
Further, in S5.4, if the standard of the model identification test is passed and reaches 95%, the identification is considered to be successful.
The invention has the beneficial effects that: the device has the advantages that lithology recognition and drilling histogram mapping can be automatically and rapidly completed, the device is simple to disassemble and install, the indoor operation time of reconnaissance personnel is saved, and the device is very suitable for field drilling core cataloging work.
Drawings
Fig. 1 is a schematic structural diagram of a core image acquisition system according to the present invention.
Detailed Description
To make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be further described with reference to the accompanying drawings.
The method mainly aims to improve the traditional core box by combining a neural network image recognition technology, set up mark points, photograph the core by a mobile phone erected above the core box, and then divide and splice the core by using built-in core image processing software. And finally, intelligently identifying and dividing each section of core to generate a drilling histogram of the whole hole.
The device has the advantages that lithology recognition and drilling histogram mapping can be automatically and rapidly completed, the device is simple to disassemble and install, the indoor operation time of reconnaissance personnel is saved, and the device is very suitable for field drilling core cataloging work.
As shown in fig. 1, a core box comprises a box body (10), the box body (10) is of a cuboid structure, the upper end of the box body is open, an object placing groove (11) of a rectangular structure is arranged in the box body, a plurality of clamping grooves (12) distributed at intervals are arranged at the upper end of one side of the box body (10), a plurality of clamping hooks (13) matched with the clamping grooves (12) are arranged at the upper end of the opposite side of the box body, and a mounting hole (14) for mounting a support (20) is formed in the box body (10).
In the invention, the side with longer length of the box body (10) is defined as a long side, and the side with shorter length is defined as a wide side. The long edge of the box body (10) is provided with scale marks, and the scale marks are used for observing the novel length of the rock core. The storage groove (11) is used for placing the rock core. Preferably, 2 clamping grooves (12) are arranged, the 2 clamping grooves (12) are respectively arranged at two ends of the upper end of the long edge of the box body (10), correspondingly, the clamping hook (13) is in an inverted L shape, and the 2 clamping grooves are also arranged at two ends of the upper end of the other long edge. The clamping grooves (12) and the clamping hooks (13) are used for realizing the connection and fixation between the two box bodies (10), namely the clamping hooks (13) are inserted into the corresponding clamping grooves (12), so that the fixation of the two box bodies (10) can be realized. In practical application, the number of the clamping grooves (12) and the clamping hooks (13) can be increased according to requirements so as to ensure the fixing effect between the two box bodies (10). The clamping groove (12) and clamping hook (13) type connecting structure has the advantages of operation connection, low implementation cost and the like. In addition, the two box bodies (10) can be connected and fixed not only through the clamping grooves (12) and the clamping hooks (13), but also through connecting structures such as bolts and buckles. The mounting hole (14) is formed in the wide edge of the box body (10) and used for mounting a support (20) of equipment for collecting the core image, and the mounting hole (14) is formed in the wide edge of the box body (10) and can be used for enabling the equipment to collect the image information of the core conveniently.
Compared with the prior art that all the rock core boxes with wood structures are wood structures, the rock core boxes are disposable and can only be used for storing rock cores. The existing invention patent only considers the convenience of storage and transportation of the rock core and can not complete the whole-hole rock core identification and cataloguing work.
The utility model provides a rock core image acquisition system, includes at least one rock core box, support (20) and image acquisition unit (30), support (20) are "L" shape mechanism, and its vertical section detachably installs in mounting hole (14), its horizontal segment extends to put the top of thing groove (11), the horizontal segment of support (20) is equipped with fixation clamp (40), fixation clamp (40) are used for fixed image acquisition unit (30).
In practical application, if only one group of rock cores need to be subjected to image acquisition, only one rock core box needs to be designed, if multiple groups of rock cores need to be subjected to image acquisition simultaneously, corresponding number of rock core boxes need to be designed, multiple rock core boxes are assembled, and a support (20) is installed in an installation hole (14) in the rock core box located at the center position to provide an optimal image acquisition visual angle and ensure the integrity of image information acquisition of multiple groups of rock cores. The bottom of the bracket (20) can be fixed by arranging a rubber ring in the mounting hole (14) or arranging a thread between the bracket (20) and the mounting hole (14). As one embodiment of the present invention, the core image acquisition system includes 5 core boxes. The image acquisition unit (30) in the invention can be equipment with an image acquisition function, such as a mobile phone, a camera and the like. The structure of the fixing clip (40) is not limited, and the structures of the fixing clip (40) which can realize the holding of the mobile phone and the camera fixing clip (40) in the prior art are all taken as specific embodiments of the fixing clip (40) in the invention. The core image acquisition system has the advantages of small volume, convenience in disassembly and assembly, convenience in transportation and the like.
An intelligent core cataloging method comprises the following steps:
s1, assembling a core image acquisition system: selecting 5 box bodies for assembling and fixing to obtain a core box group, fixing a support in a mounting hole of the box body positioned at the central position, and mounting an image acquisition unit on a fixing clamp; debugging the image acquisition distance of the image acquisition unit, and placing a target rock core in each box body; specifically, the image acquisition unit is a mobile phone. Optimally, to facilitate later pre-processing of the borehole core image, the color of the core box may be designed to be a color that is clearly distinct from the borehole core color, such as red. When the identification is carried out under the condition that the shooting environment light is dim, the mobile phone flashlight is required to be turned on to carry out light supplement, so that the identification result is not influenced by chromatic aberration.
S2, camera correction: marking points are arranged at four right angles of the box body of each core box, the focal length of the image acquisition unit is adjusted and debugged, so that the whole view field covers the core box, and meanwhile, the image information of the acquired core box is clear. Specifically, the focal length is automatically adjusted through the mobile phone, the four mark points are aligned, the whole view field covers the core box, and meanwhile, the shot pictures of the core box are clear and pollution-free. And the sizes of the shot photos are marked through the scales on the edges of the rock core box body, so that the length of each section of lithology is conveniently identified.
S3, image acquisition: after the core image acquisition system is assembled, shooting the whole core box group by using a mobile phone, and acquiring the core box group image;
s4, image preprocessing:
cutting and dividing the collected core box group images to obtain columnar images of single drill hole cores, splicing 5 columnar images of the single drill hole cores, randomly intercepting the spliced drill hole core columnar images according to a preset pixel size, preprocessing all intercepted images, and then zooming to a preset size to obtain preprocessed images with unified specifications; specifically, the method comprises the following steps:
s4.1, cutting the core box image by using an image segmentation technology, segmenting the core box image into five histograms of single drill cores, and then splicing;
s4.2, randomly intercepting the spliced drill core histogram according to the size of 320 multiplied by 320 pixels;
s4.3, turning and rotating the randomly captured picture;
s4.4, fine adjustment of the brightness of the drill hole core image obtained through processing is carried out;
and S4.5, uniformly scaling all the obtained pictures to 256 multiplied by 256 pixels.
It should be noted that the image preprocessing techniques designed in this step, such as image cropping, cutting, screenshot, flipping, rotating, brightness adjustment, and pixel scaling, all belong to the prior art.
S5, establishing an image recognition model:
establishing a drilling core image recognition model by using a deep learning algorithm based on a convolutional neural network and by using an Inception V3 pre-training model method, and performing model training by using the drilling core image subjected to image preprocessing in S4;
specifically, the method comprises the following steps:
s5.1, preparing data: collecting known types of drilling core images, processing the known types of drilling core images according to the method in the step 4, labeling each drilling core image, and establishing a drilling core model training set;
s5.2, model construction: completely extracting the convolutional layer and the pooling layer of the convolutional neural network Inception V3, removing the last output layer of the Inception V3 model, adding two initialized full-connected layers, and extracting the image characteristics of the drill core by adopting the parameter values of the Inception V3 model;
s5.3, model training: training the drill core recognition model by adopting the training sample obtained in the S4.5;
s5.4, model testing: and (4) pushing the known borehole core images except the training set to preprocessing according to the method in the step 4, establishing a borehole core test set, and checking the identification accuracy of the model by using the test set. And if the model can be correctly identified, the model is qualified, if the model cannot be correctly identified, the training sample is collected again to train the model and adjust the parameters of the model until the model can be correctly identified, and specifically, if the standard of passing the model identification test reaches 95%, the identification is considered to be successful.
It should be noted that the establishment of the image recognition model in the present invention belongs to the prior art, and for example, the technical scheme of step S5 in the present invention is described in the text of the lithology automatic recognition and classification method based on rock image deep learning (zhangye, lymingtao, hanshuai, 2018, lithology automatic recognition and classification method based on rock image deep learning, 34 (2): 333-342), and therefore, the detailed operation process of the present invention is not described again.
S6, image recognition:
and S5, after the training of the drill hole core image recognition model is finished, the method in the step 4 is used for obtaining a preprocessed drill hole core image, namely the drill hole core image to be recognized is used as a drill hole core image to be recognized and is guided into the recognition model for recognition.
In practical application, the trained deep learning model can be installed in the intelligent recognition program of the drill core as a built-in file, so that the operation of training in advance during recognition is avoided, and the recognition function is directly realized.
S7, returning an identification result:
and after S6, respectively obtaining the core type and the core depth information of each section of core column according to the recognition result, marking the core depth corresponding to the core type, and establishing a core column CAD graph.
It should be noted that, in the present invention, a core cylindrical CAD graph is derived according to information such as the type and depth of an acquired core, which is a prior art, for example, the related technical principles are described in the patent application No. 202111509800.2 entitled method for reconstructing a surrounding rock fracture in a three-dimensional simulation test, an electronic device, and a storage medium, which is a conventional technical means in the field, and belongs to the prior art, and the present invention does not explain a specific operation method and principle thereof.
In this document, the terms front, back, upper and lower are used to define the positions of the components in the drawings and each other, and are used for the sake of clarity and convenience in technical solution. It is to be understood that the use of the directional terms should not be taken to limit the scope of the claims.
The features of the embodiments and embodiments described herein above may be combined with each other without conflict.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (8)
1. The utility model provides a core box, its characterized in that, includes the box body, the box body is the cuboid structure, and open setting in its upper end is equipped with the thing groove of putting of rectangle structure in it, one side upper end of box body is equipped with a plurality of interval distribution draw-in grooves, and the upper end of its relative one side is equipped with a plurality of the pothook of draw-in groove looks adaptation, be equipped with the mounting hole that is used for the installing support on the box body.
2. A core image acquisition system, characterized by, includes support, image acquisition unit and at least one core box as in claim 1, the support is "L" shape mechanism, and its vertical section detachably installs in the mounting hole, its horizontal segment extends to put the top of thing groove, the horizontal segment of support is equipped with the fixation clamp, the fixation clamp is used for fixed image acquisition unit, image acquisition unit is used for gathering the core image.
3. The core image acquisition system as claimed in claim 2, wherein the image acquisition unit is a mobile phone.
4. An intelligent logging method for a rock core, which is implemented by using the rock core image acquisition system of claim 2 or 3, and comprises the following steps:
s1, assembling a core image acquisition system, and placing a drill core in a core box;
s2, camera correction: setting mark points at four right angles of the box body of each core box, adjusting the focal length of the debugging image acquisition unit to enable the whole view field to cover the core box, and simultaneously ensuring that the acquired image information of the core box is clear;
s3, after the core image acquisition system is assembled, shooting a core box by using the image acquisition unit, and acquiring an image of the core box, wherein if the core image acquisition system comprises a plurality of core boxes, an image of a core box group consisting of the plurality of core boxes is acquired;
s4, image preprocessing: cutting and dividing the collected core box images or core box group images to obtain columnar images of single drill cores, and preprocessing the columnar images of the single drill cores to obtain preprocessed pictures with unified specifications;
s5, establishing an image recognition model: establishing a drilling core image recognition model by using a deep learning algorithm based on a convolutional neural network and by using an Inception V3 pre-training model method, and performing model training by using the drilling core image subjected to image preprocessing in S4;
s6, after the training of the drilling core image recognition model is finished in S5, selecting the picture processed in S4 as the drilling core image to be recognized, and guiding the picture into the recognition model for recognition;
and S7, after the image identification is completed in the S6, respectively obtaining the core type and the core depth information of each section of core column according to the identification result, marking the core depth corresponding to the core type, and establishing a core column CAD graph.
5. The intelligent core cataloging method according to claim 4, wherein in S1, if the core image collecting system comprises a plurality of core boxes, the plurality of core boxes are assembled and fixed to obtain a core box group, the bracket is fixed in a mounting hole of a box body positioned at a central position, and the image collecting unit is mounted on the fixing clamp; and debugging the image acquisition distance of the image acquisition unit, and placing a target rock core in each box body.
6. The intelligent core cataloging method of claim 4, wherein the image preprocessing in S4 comprises the following specific steps:
s4.1, cutting the core box image or the core box group image by using an image segmentation technology, segmenting the core box image or the core box group image into a histogram of a single drilling core, and splicing the obtained histograms of a plurality of drilling cores if the number of core boxes is multiple;
s4.2, randomly intercepting the single or spliced drill core histogram according to the size of 320 multiplied by 320 pixels;
s4.3, turning and rotating the randomly captured picture;
s4.4, fine adjustment of the brightness of the drill hole core image obtained through processing is carried out;
and S4.5, uniformly scaling all the obtained pictures to 256 multiplied by 256 pixels to obtain the preprocessed drilling core image.
7. The intelligent core cataloging method of claim 4, wherein the image recognition model in S5 comprises the following specific steps:
s5.1, preparing data: collecting drilling core images of known types, processing the drilling core images of the known types according to the method in the step 4, labeling each drilling core image, and establishing a drilling core model training set;
s5.2, model construction: completely extracting the convolutional layer and the pooling layer of the convolutional neural network Inception V3, removing the last output layer of the Inception V3 model, adding two initialized full-connected layers, and extracting the image characteristics of the drill core by adopting the parameter values of the Inception V3 model;
s5.3, model training: training the drill core recognition model by adopting the training sample obtained in the step S4;
s5.4, model testing: pushing the known borehole core images except the training set to preprocessing according to the method in the step 4, establishing a borehole core test set, and inspecting the identification accuracy of the model by using the test set; if the model can be identified correctly, the model is qualified, if the model is not qualified, the training sample is collected again to train the model and adjust the parameters of the model until the model can be identified correctly.
8. The intelligent core cataloging method of claim 7, wherein in S5.4, if the model identification test passes the standard of 95%, the identification is considered to be successful.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211474459.6A CN115753612A (en) | 2022-11-22 | 2022-11-22 | Intelligent core box, core image acquisition system and intelligent core recording method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211474459.6A CN115753612A (en) | 2022-11-22 | 2022-11-22 | Intelligent core box, core image acquisition system and intelligent core recording method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115753612A true CN115753612A (en) | 2023-03-07 |
Family
ID=85335944
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211474459.6A Pending CN115753612A (en) | 2022-11-22 | 2022-11-22 | Intelligent core box, core image acquisition system and intelligent core recording method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115753612A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116309807A (en) * | 2023-04-04 | 2023-06-23 | 中南大学 | Core ground repositioning system based on digital image intelligent recognition |
-
2022
- 2022-11-22 CN CN202211474459.6A patent/CN115753612A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116309807A (en) * | 2023-04-04 | 2023-06-23 | 中南大学 | Core ground repositioning system based on digital image intelligent recognition |
| CN116309807B (en) * | 2023-04-04 | 2024-04-26 | 中南大学 | Core ground repositioning system based on digital image intelligent recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110378879A (en) | A kind of Bridge Crack detection method | |
| CN112241711B (en) | Intelligent method for identifying RQD from drilling core photo | |
| CN109887020A (en) | A kind of plant organ's separation method and system | |
| CN104036235A (en) | Plant species identification method based on leaf HOG features and intelligent terminal platform | |
| CN108109157B (en) | Rock mass evaluation analysis method based on digital panoramic borehole image | |
| CN107704878A (en) | A kind of high-spectral data storehouse semi-automation method for building up based on deep learning | |
| CN115753612A (en) | Intelligent core box, core image acquisition system and intelligent core recording method | |
| CN110414341A (en) | A kind of Blasting In The Open Mining heap measurement statistical method based on unmanned air vehicle technique | |
| CN110837839B (en) | High-precision unmanned aerial vehicle orthographic image manufacturing and data acquisition method | |
| CN110688955A (en) | Building construction target detection method based on YOLO neural network | |
| CN108509993A (en) | A kind of water bursting in mine laser-induced fluorescence spectroscopy image-recognizing method | |
| CN113743358B (en) | Landscape vision feature recognition method adopting omnibearing collection and intelligent calculation | |
| CN110619297B (en) | Bean fruiting body image batch acquisition and recognition method and device | |
| CN109146880A (en) | A kind of electric device maintenance method based on deep learning | |
| CN110648366A (en) | Orange detection system and orange sectioning method based on deep learning | |
| CN114067094A (en) | Multispectral camera system for rock drilling jumbo and intelligent surrounding rock identification method | |
| CN114663787A (en) | Single-tree segmentation method fusing unmanned aerial vehicle CHM and RGB images | |
| CN106969753B (en) | Unmanned plane data processing method based on Electric Design application | |
| CN115035297A (en) | Automatic recording method, system, device and medium for drilling core RQD | |
| CN112966698A (en) | Freshwater fish image real-time identification method based on lightweight convolutional network | |
| CN112483071A (en) | Drilling core image recognition device and method | |
| CN112001964A (en) | Flood evolution process inundation range measuring method based on deep learning | |
| CN114138014B (en) | Unmanned aerial vehicle control method, device, equipment and storage medium for land survey | |
| CN112287787B (en) | Crop lodging grading method based on gradient histogram characteristics | |
| CN212413758U (en) | Phenotype measuring instrument for single plant test of crops |
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 |