CN116754609B - Method and device for detecting rust of chilled water pipe of data center - Google Patents
Method and device for detecting rust of chilled water pipe of data center Download PDFInfo
- Publication number
- CN116754609B CN116754609B CN202311047384.8A CN202311047384A CN116754609B CN 116754609 B CN116754609 B CN 116754609B CN 202311047384 A CN202311047384 A CN 202311047384A CN 116754609 B CN116754609 B CN 116754609B
- Authority
- CN
- China
- Prior art keywords
- rust
- image
- layer
- water pipe
- laser
- 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
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 75
- JEIPFZHSYJVQDO-UHFFFAOYSA-N iron(III) oxide Inorganic materials O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 title claims abstract description 74
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000001514 detection method Methods 0.000 claims abstract description 75
- 230000004927 fusion Effects 0.000 claims abstract description 50
- 238000012545 processing Methods 0.000 claims abstract description 45
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 22
- 238000012937 correction Methods 0.000 claims abstract description 17
- 238000006073 displacement reaction Methods 0.000 claims abstract description 17
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 10
- 230000007613 environmental effect Effects 0.000 claims abstract description 9
- 238000012512 characterization method Methods 0.000 claims abstract description 6
- 238000005070 sampling Methods 0.000 claims abstract description 4
- 238000010606 normalization Methods 0.000 claims description 21
- 238000004321 preservation Methods 0.000 claims description 20
- 238000007710 freezing Methods 0.000 claims description 19
- 230000008014 freezing Effects 0.000 claims description 17
- 230000004913 activation Effects 0.000 claims description 15
- 230000005540 biological transmission Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 12
- 238000007499 fusion processing Methods 0.000 claims description 10
- 238000005507 spraying Methods 0.000 claims description 10
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 4
- 238000009413 insulation Methods 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 230000005855 radiation Effects 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 22
- 239000012774 insulation material Substances 0.000 description 8
- 230000007797 corrosion Effects 0.000 description 4
- 238000005260 corrosion Methods 0.000 description 4
- 238000001816 cooling Methods 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 229920000742 Cotton Polymers 0.000 description 2
- 239000011248 coating agent Substances 0.000 description 2
- 238000000576 coating method Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011049 filling Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 239000004793 Polystyrene Substances 0.000 description 1
- 229920005830 Polyurethane Foam Polymers 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003518 caustics Substances 0.000 description 1
- 238000005253 cladding Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000011888 foil Substances 0.000 description 1
- 239000011491 glass wool Substances 0.000 description 1
- 229920001903 high density polyethylene Polymers 0.000 description 1
- 239000004700 high-density polyethylene Substances 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 239000006247 magnetic powder Substances 0.000 description 1
- 238000011418 maintenance treatment Methods 0.000 description 1
- 239000011490 mineral wool Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229920002223 polystyrene Polymers 0.000 description 1
- 239000011496 polyurethane foam Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N25/00—Investigating or analyzing materials by the use of thermal means
- G01N25/72—Investigating presence of flaws
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J11/00—Manipulators not otherwise provided for
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/32—Indexing scheme for image data processing or generation, in general involving image mosaicing
-
- 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
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- 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
- G06T2207/10048—Infrared 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
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Immunology (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Pathology (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Quality & Reliability (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The application discloses a method and a device for detecting rust of a chilled water pipe of a data center, wherein the method comprises the steps that an infrared detection part samples the chilled water pipe coated with an insulating layer according to a preset sampling frequency; the laser detection part receives the laser signals of the environmental object and the target laser signals and generates an accurate three-dimensional point cloud picture; fusion correction is carried out on the thermal infrared spliced image based on the positioning mark, the three-dimensional point cloud image and the preset interval displacement parameter, so that a fusion image is formed; and (3) carrying out data processing on the fusion image by combining empirical mode decomposition with histogram equalization, inputting the processed fusion image into a convolutional neural network for rust detection, and classifying the rust degree according to rust characterization. By adopting the embodiment of the application, on one hand, the rust detection precision can be improved, and on the other hand, the accurate marking of the rust area and the rust degree is facilitated, and the treatment efficiency of subsequent maintenance is improved.
Description
Technical Field
The application relates to the technical field of data center maintenance, in particular to a method and a device for detecting rust of a chilled water pipe of a data center.
Background
Along with the information becoming the major trend of economic and social development in the world today, the rapidly growing data volume makes the original storage capacity and application system difficult to meet urgent needs, so that it is necessary to construct a high-reliability and large-capacity large data center, the high-reliability operation of the data center is derived from the safe maintenance of the data center, the data center chilled water pipe is one of pipeline systems for carrying out the cooling of the service machine room by adopting a water cooling unit and the data center to transport chilled water to the service machine room precise air conditioner through a cold source, namely the water cooling unit, and is an important facility for the safe maintenance of the data center.
Because the chilled water pipe of data center is because of the temperature is too low, in order to prevent the comdenstion water and prolong the consideration of pipeline life, adopt insulation material to carry out the cladding in order to overcome the defect, but along with insulation material life's growth, the problem that early construction stage insulation material parcel quality is low and not in place, lead to pipeline insulation material to drop and not laminate, the pipeline surface contacts with the air, lead to the pipeline to produce corrosion, and the pipeline is covered by insulation material, patrol inspector is difficult to directly discover pipeline corrosion through naked eyes, greatly harm data center's safety.
In the prior art, pipeline corrosion detection comprises ultrasonic detection, magnetic powder detection, magnetic memory detection, laser scanning technology, pipeline inner wall corrosion detection technology, X-ray detection technology, thermal infrared image detection and the like, but the direct application of the detection method to a data center has the difficulty that firstly, the data center usually operates all the day and cannot be subjected to electromagnetic interference, secondly, a frozen water pipeline wraps heat preservation cotton, detection workload is too large after the heat preservation cotton is disassembled, thirdly, the space of the data center is limited, large detection equipment is difficult to place, fourthly, the existing detection mode only considers the detection result of the current position, the detection result of the current position is positioned inaccurately, the detection result is not in butt joint with absolute space reference, and the degree and the result of the rust detection are not favorable to be transmitted to maintenance personnel for automatic maintenance treatment.
In view of the foregoing, a scheme is needed to make the detection of pipeline rust in a data center more accurate and improve the operation safety of the data center.
Disclosure of Invention
In view of the above, the application provides a method and a device for detecting rust of a chilled water pipe of a data center.
The technical scheme of the application is realized as follows:
in a first aspect, a device for detecting rust in a chilled water line of a data center, characterized in that,
the device comprises a stepping part, a marking part, a laser detection part, an infrared detection part, an environment reference part and an external processing host terminal;
the stepping part comprises a stepping device sleeved on a heat preservation layer of a chilled water pipe, the stepping device comprises a semi-annular body and a component support, the semi-annular body comprises a fixed bridge in the middle of the body and two stretchable elastic ring arms, one end of the fixed bridge, which is away from the heat preservation layer of the chilled water pipe, is connected with the component support, the end parts of the two stretchable elastic ring arms of the semi-annular body and one end of the fixed bridge, which is close to the heat preservation layer of the chilled water pipe, are respectively connected with an electric hydraulic connecting rod, the electric hydraulic connecting rod connected with the fixed bridge is provided with a driving sliding wheel, a remotely adjustable stepping motor is arranged on the driving sliding wheel, the electric hydraulic corresponding to the end parts of the two stretchable elastic ring arms are connected with driven wheels, and when the stepping motor is started, the driving sliding wheel drives the driven wheels to advance along the trend direction of the chilled water pipe, which is clung to the heat preservation layer of the chilled water pipe;
the marking part is arranged on at least one driven wheel and comprises a spraying device and an encoder, wherein the encoder is used for measuring relative displacement, and the spraying device is used for spraying a plurality of positioning marks on the insulating layer of the chilled water pipe at intervals according to preset interval displacement parameters;
the laser detection part is positioned on the component bracket and comprises a laser emitter, a rotary scanner, a laser receiver, a clock, a counter, a data processor and a first data wireless transmission module;
the infrared detection part is positioned on the component bracket and comprises an infrared camera lens, an infrared sensing array, a data acquisition card and a second data wireless transmission module;
the environmental benchmark section comprises a target which is a known benchmark pre-arranged near a chilled water pipe of a data center;
the external processing host terminal comprises a three-dimensional point cloud image characteristic processing unit, a thermal infrared characteristic processing unit, a fusion processing unit and a rust detection unit; the three-dimensional point cloud image feature processing unit sets a first feature recognition operator according to the positioning mark, the first feature recognition operator recognizes the positioning mark, and then determines the absolute coordinate of the positioning mark in the three-dimensional point cloud image by combining with the target; the thermal infrared characteristic processing unit sets a second characteristic recognition operator according to the positioning marks, the second characteristic recognition operator recognizes the positioning marks, when at least two positioning marks which are arranged at intervals are scanned and recognized on a thermal infrared image, the image pixel positions of the positioning marks in the thermal infrared image are determined, when at least one positioning mark is recognized by an adjacent thermal infrared image, the adjacent thermal infrared images are spliced to form a thermal infrared spliced image, and the fusion processing unit carries out fusion correction with the thermal infrared spliced image based on the positioning marks, the three-dimensional point cloud image and preset interval displacement parameters to form a fusion image; and the rust detection unit constructs a convolutional neural network to process the fusion image obtained by the fusion processing unit so as to obtain a rust detection result.
As a further alternative, the laser detection unit includes a laser transmitter, a rotary scanner, a laser receiver, a clock and a counter, a data processor, and a first data wireless transmission module, where the laser transmitter transmits a laser beam, the rotary scanner scans the environment around the whole chilled water pipe in horizontal and vertical directions, the laser receiver receives the laser beam and performs photoelectric conversion, the clock and the counter are used to measure the round trip time of the laser beam and calculate the distance from an object in the environment to the laser radar, the data processor is used to process the received laser signal of the environmental object and the target laser signal, and generate an accurate three-dimensional point cloud image, and the first data wireless transmission module sends the three-dimensional point cloud image to the external processing host terminal;
as a further alternative, the infrared camera lens is disposed on one side of the component support near the pipeline, the infrared sensing array receives heat radiation signals of the chilled water pipe acquired by the infrared camera lens, the data acquisition card is used for photoelectric conversion, and the second data wireless transmission module sends a plurality of thermal infrared images acquired by the photoelectric conversion to the external processing host terminal.
As a further alternative, the first feature recognition operator recognizes the positioning mark, and combining the target, and then determining absolute coordinates of the positioning mark in the three-dimensional point cloud image includes: if more than two positioning marks are identified, vector connection is carried out between the positioning marks and the targets, adjustment is carried out by taking the preset interval displacement parameter as a distance constraint condition, an absolute coordinate correction value of the positioning marks in the three-dimensional point cloud picture is obtained according to the adjustment, and the absolute coordinates of the positioning marks in the three-dimensional point cloud picture are determined based on the absolute coordinate correction value.
As a further alternative, the positioning mark is characterized in that it has a serial number coding function.
In a second aspect, a rust detecting method for detecting a rust device of a chilled water pipe of a data center by using the foregoing method is characterized in that:
step one: starting a stepping motor, wherein the driving sliding wheel drives the driven wheel to be clung to the insulating layer of the frozen water pipe to advance along the trend direction of the frozen water pipe; the infrared detection part samples the frozen water pipe coated with the heat preservation layer according to a preset sampling frequency; the laser detection part receives the laser signals of the environmental object and the target laser signals and generates an accurate three-dimensional point cloud picture; fusion correction is carried out on the thermal infrared spliced image based on the positioning mark, the three-dimensional point cloud image and the preset interval displacement parameter, so that a fusion image is formed;
secondly, carrying out graying and normalization operations on the fusion image, constructing a gaussian filter, filtering the fusion image, and carrying out data processing by a method of empirical mode decomposition combined with histogram equalization, wherein the specific data processing process is to decompose the fusion image into IMF1 to IMF6 through empirical mode decomposition, wherein the IMF1 has the highest frequency and the IMF6 has the lowest frequency, carrying out histogram equalization processing on IMF2-IMF6 components, setting a preset threshold value for IMF1 components, filtering high-frequency noise, and carrying out reconstruction and normalization on the IMF1-IMF6 to obtain a processed fusion image;
inputting the processed fusion image into a convolutional neural network for rust detection, classifying the rust degree according to rust characterization, wherein the convolutional neural network is established by acquiring a data center chilled water pipe infrared rust image labeling set as a test set, adopting the fusion image as a training set, training the convolutional neural network through the test set and the training set, and obtaining a rust detection result by utilizing the trained convolutional neural network.
As a further alternative, the convolutional neural network comprises an input layer, a convolutional layer and a pooling layer, a batch normalization layer and an activation function layer, a fully connected layer design, and model optimization.
As a further alternative, the input layer is characterized by: assume that the size of the input fusion image is w×h, where W and H represent the width and height of the image, respectively;
convolution layer and pooling layer: extracting local features by using a smaller convolution kernel size, gradually extracting features of higher layers by using a plurality of convolution layers, and then performing dimension reduction processing on the features by using a pooling layer;
batch normalization layer and activation function layer: adding a batch normalization layer and an activation function layer between the convolution layer and the full connection layer, using a ReLU as an activation function, and adding the batch normalization layer to normalize the output;
full tie layer: using a full-connection layer with a plurality of neurons as an output layer and using a softmax function as an activation function to generate probability values of each category, representing the probability that the image belongs to different categories, defining 4 categories, and respectively corresponding to no-rust, slight-rust, moderate-rust and serious-rust, wherein the formula of the full-connection layer is as follows:
y = softmax(W*x + b)
wherein x represents an input feature, W and b represent weights and bias parameters of the full-connection layer, and y represents a probability value of each category;
model optimization: the cross entropy is used as a loss function, a random gradient descent algorithm is used for training and optimizing the model, and a regularization technology and a dropout technology are adopted to avoid overfitting, wherein the formula of the cross entropy is as follows:
Loss = -1/N * sum(Z*log(y_hat))
where N represents the number of samples, Z represents the true label, and y_hat represents the probability value predicted by the model.
In a third aspect, a stepper robot is provided for implementing the method described above.
In a fourth aspect, a computer storage medium has stored thereon a computer program which, when executed by a processor, implements the method described above.
The application provides a method and a device for detecting rust of a chilled water pipe of a data center and a storage medium thereof, which at least have the following beneficial effects:
1. through setting up step-by-step portion, when step motor starts, initiative movable pulley drives from the driving wheel hugs closely the direction of freezing water pipe heat preservation along the direction of freezing water pipe trend advances, can carry out automated operation along freezing water pipe heat preservation through above-mentioned mode, improves detection efficiency, simultaneously, avoids artifical handheld detection error.
2. The stepping part is provided with an extensible elastic ring arm and is electrically and hydraulically connected, so that the size of the freezing pipeline with different coating heat preservation layers can be better adapted and attached.
3. The method is characterized in that the method comprises the steps of providing a relative sequence number coding function to realize accurate labeling, transmitting an absolute space reference standard, and taking the absolute space reference standard as an image processing characteristic and participating in error correction.
4. Acquiring a three-dimensional point cloud image and a thermal infrared image, performing target processing to obtain the three-dimensional point cloud image containing a positioning mark at an accurate position, firstly splicing the thermal infrared image according to the mark, and then performing fusion correction on the fusion processing unit and the thermal infrared spliced image based on the positioning mark, the three-dimensional point cloud image and a preset interval displacement parameter to form a fusion image; and (3) carrying out graying and normalization operations on the fusion image, constructing a gaussian filter to filter the fusion image, carrying out data processing by a method of empirical mode decomposition and histogram equalization, then inputting the processed fusion image into a convolutional neural network to carry out rust detection, and classifying the rust degree according to rust characterization.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a device for detecting rust in a chilled water tube of a data center according to the present application.
In the figure, a 1-freezing water pipe, a 2-freezing water pipe heat insulation layer, a 3-semi-annular body, a 4-component bracket, a 5-fixed bridge, a 6-elastic ring arm, a 7-electro-hydraulic connecting rod, an 8-driving sliding wheel, a 9-driven wheel, a 10-spraying device, an 11-laser detection part and a 12-infrared detection part.
Detailed Description
The following description of the technical solutions in the embodiments of the present application will be clear and complete, and it is obvious that the described embodiments are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Referring to fig. 1, in a first aspect, a device for detecting rust in a chilled water line of a data center, comprising: the device comprises a stepping part, a marking part, a laser detection part, an infrared detection part, an environment reference part and an external processing host terminal;
the stepping part comprises a stepping device sleeved on the insulating layer 2 of the freezing water pipe, the stepping device comprises a semi-annular body 3 and a component support 4, the semi-annular body comprises a fixed bridge 5 in the middle of the body and two extensible elastic ring arms 6, one ends of the fixed bridge, which deviate from the insulating layer of the freezing water pipe, are connected with the component support 4, the end parts of the two extensible elastic ring arms of the semi-annular body and one end of the fixed bridge, which is close to the insulating layer of the freezing water pipe, are respectively connected with an electric hydraulic connecting rod 7, the electric hydraulic connecting rods connected with the fixed bridge are provided with an active sliding wheel 8, the electric hydraulic connecting rods corresponding to the end parts of the two extensible elastic ring arms are connected with driven wheels 9, and when the stepping motor is started, the active sliding wheel 8 drives the driven wheels 9 to be tightly attached to the insulating layer of the freezing water pipe and move forward along the direction of the freezing water pipe 1;
it will be appreciated that the remotely adjustable motor may be a motor that controls the rotational speed of the motor or stops by remote control, which may be controlled by wireless remote control or wired network, the application is not further limited to a particular type of motor.
Generally, the insulation layer of the chilled water pipe comprises an outer protection layer, an insulation material filling layer and an inner liner layer, wherein the outer protection layer is used for protecting the pipe insulation layer from rainwater, corrosive substances and fire; the heat insulation material filling layer is used for reducing energy loss in the pipeline conveying process, and the material types comprise polyurethane foam, polystyrene board, glass wool, rock wool and the like; the lining layer is used for preventing the heat insulation material from being corroded and damaged by the medium, and is laid by adopting high-density polyethylene, aluminum foil and other materials.
It is worth mentioning that, through setting up step-by-step portion, when step motor starts, initiative movable pulley drives from the driving wheel and hugs closely the direction that the freezing water pipe heat preservation was followed to the freezing water pipe trend advances, can carry out automated operation along the freezing water pipe heat preservation through above-mentioned mode, improves detection efficiency, simultaneously, avoids artifical handheld detection error.
In addition, the step part is provided with an extensible elastic ring arm and an electric hydraulic connection, so that the size of the freezing pipeline with different coating heat preservation layers can be better adapted and attached.
The marking part is arranged on at least one driven wheel 9 and comprises a spraying device 10 and an encoder, wherein the encoder is used for measuring relative displacement, and the spraying device 10 is used for spraying a plurality of positioning marks on the frozen water pipe heat preservation layer 2 at intervals according to preset interval displacement parameters;
the laser detection part 11 is positioned on the component bracket 4 and comprises a laser emitter, a rotary scanner, a laser receiver, a clock and counter, a data processor and a first data wireless transmission module;
the infrared detection part 12 is positioned on the component bracket 4 and comprises an infrared camera lens, an infrared sensing array, a data acquisition card and a second data wireless transmission module;
the environmental reference section includes a target, which is a known reference point arranged in advance in the vicinity of the chilled water tube 1 of the data center;
the external processing host terminal comprises a three-dimensional point cloud image characteristic processing unit, a thermal infrared characteristic processing unit, a fusion processing unit and a rust detection unit; the three-dimensional point cloud image feature processing unit sets a first feature recognition operator according to the positioning mark, the first feature recognition operator recognizes the positioning mark, and then determines the absolute coordinate of the positioning mark in the three-dimensional point cloud image by combining with the target; the thermal infrared characteristic processing unit sets a second characteristic recognition operator according to the positioning marks, the second characteristic recognition operator recognizes the positioning marks, when at least two positioning marks which are arranged at intervals are scanned and recognized on a thermal infrared image, the image pixel positions of the positioning marks in the thermal infrared image are determined, when at least one positioning mark is recognized by an adjacent thermal infrared image, the adjacent thermal infrared images are spliced to form a thermal infrared spliced image, and the fusion processing unit carries out fusion correction with the thermal infrared spliced image based on the positioning marks, the three-dimensional point cloud image and preset interval displacement parameters to form a fusion image; and the rust detection unit constructs a convolutional neural network to process the fusion image obtained by the fusion processing unit so as to obtain a rust detection result.
It should be noted that, for the existing detection method, only the detection result of the current position is considered, the detection result of the current position is positioned inaccurately, and the detection result is not abutted against the absolute space reference, therefore, the marking part and the environment reference part are provided, and the functions have the following effects that one relative sequence number coding function is provided, the absolute space reference is transmitted, and the third is taken as the image processing feature and participates in error correction.
It will be appreciated that the chilled water pipe rusting device is affected by the viewing angle of the laser or the infrared sensor and has an observation blind area, so that after the rusting device performs a round of forward automatic collection along the advancing first direction of the water pipe, the rusting device can be arranged on the opposite side of the water pipe according to the positioning mark, so that the rusting device can perform a round of reverse automatic collection along the opposite second direction of the water pipe, and the observation blind area is eliminated.
It is worth to say that the automatic collection of the forward direction and the reverse direction also needs to be fused and corrected by means of the positioning mark so as to realize accurate rust detection.
As a further alternative, the laser detection unit includes a laser transmitter, a rotary scanner, a laser receiver, a clock and a counter, a data processor, and a first data wireless transmission module, where the laser transmitter transmits a laser beam, the rotary scanner scans the environment around the whole chilled water pipe in horizontal and vertical directions, the laser receiver receives the laser beam and performs photoelectric conversion, the clock and the counter are used to measure the round trip time of the laser beam and calculate the distance from an object in the environment to the laser radar, the data processor is used to process the received laser signal of the environmental object and the target laser signal, and generate an accurate three-dimensional point cloud image, and the first data wireless transmission module sends the three-dimensional point cloud image to the external processing host terminal;
as a further alternative, the infrared camera lens is disposed on one side of the component support near the pipeline, the infrared sensing array receives heat radiation signals of the chilled water pipe acquired by the infrared camera lens, the data acquisition card is used for photoelectric conversion, and the second data wireless transmission module sends a plurality of thermal infrared images acquired by the photoelectric conversion to the external processing host terminal.
As a further alternative, the first feature recognition operator recognizes the positioning mark, and combining the target, and then determining absolute coordinates of the positioning mark in the three-dimensional point cloud image includes: if more than two positioning marks are identified, vector connection is carried out between the positioning marks and the targets, adjustment is carried out by taking the preset interval displacement parameter as a distance constraint condition, an absolute coordinate correction value of the positioning marks in the three-dimensional point cloud picture is obtained according to the adjustment, and the absolute coordinates of the positioning marks in the three-dimensional point cloud picture are determined based on the absolute coordinate correction value.
It should be noted that the above correction process for identifying more than two positioning marks can be performed continuously and iteratively, and preset conditions can be set to perform iterative adjustment calculation, for example, preset duration, preset step length, preset observation times, and the like, so that the accuracy of the absolute coordinates of the positioning marks in the three-dimensional point cloud image is ensured.
It will be appreciated that the feature recognition operator may identify representative feature points or feature descriptors by calculating the intensity, gradient, angle, etc. of the image, so as to implement feature extraction and matching of the image, which is generally included in Harris, SIFT (Scale-Invariant Feature Transform), SURF (Speeded Up Robust Features), FAST (Features from Accelerated Segment Test), HOG (Histogram of Oriented Gradients), etc., which is not further limited by the present application.
As a further alternative, the positioning mark is characterized in that it has a serial number coding function.
In a second aspect, a rust detecting method for detecting a rust device of a chilled water pipe of a data center by using the foregoing method is characterized in that:
step one: starting a stepping motor, wherein the driving sliding wheel drives the driven wheel to be clung to the insulating layer of the frozen water pipe to advance along the trend direction of the frozen water pipe; the infrared detection part samples the frozen water pipe coated with the heat preservation layer according to a preset sampling frequency; the laser detection part receives the laser signals of the environmental object and the target laser signals and generates an accurate three-dimensional point cloud picture; fusion correction is carried out on the thermal infrared spliced image based on the positioning mark, the three-dimensional point cloud image and the preset interval displacement parameter, so that a fusion image is formed;
secondly, carrying out graying and normalization operations on the fusion image, constructing a gaussian filter, filtering the fusion image, and carrying out data processing by a method of empirical mode decomposition combined with histogram equalization, wherein the specific data processing process is to decompose the fusion image into IMF1 to IMF6 through empirical mode decomposition, wherein the IMF1 has the highest frequency and the IMF6 has the lowest frequency, carrying out histogram equalization processing on IMF2-IMF6 components, setting a preset threshold value for IMF1 components, filtering high-frequency noise, and carrying out reconstruction and normalization on the IMF1-IMF6 to obtain a processed fusion image;
inputting the processed fusion image into a convolutional neural network for rust detection, classifying the rust degree according to rust characterization, wherein the convolutional neural network is established by acquiring a data center chilled water pipe infrared rust image labeling set as a test set, adopting the fusion image as a training set, training the convolutional neural network through the test set and the training set, and obtaining a rust detection result by utilizing the trained convolutional neural network.
It is worth mentioning that the fusion image is adopted to perform graying and normalization operations, a gaussian filter is constructed to filter the fusion image, data processing is performed through a method of empirical mode decomposition and histogram equalization, then the processed fusion image is input into a convolutional neural network to perform rust detection, and the rust degree is classified according to rust characterization, so that the accuracy of rust detection and the accuracy of rust degree classification can be effectively improved in the processing process.
It will be appreciated that the specific steps of histogram equalization include:
and counting the number of pixels appearing in each gray level in the image to obtain a gray level distribution histogram of the original image.
A cumulative distribution function (Cumulative Distribution Function, CDF) for each gray level is calculated from the gray level distribution histogram, i.e., the number of pixels for each gray level is accumulated and normalized.
A Mapping Function (Mapping Function) of each gray level is calculated according to the CDF, that is, each gray level in the original image is mapped to a new gray level, so that the new gray level is distributed as uniformly as possible.
And converting the gray level of each pixel in the original image according to the mapping function to obtain an equalized image.
Histogram equalization may also introduce some side effects, such as enhanced noise and detail, among others.
In order to improve the identification effectiveness of the rust detection convolutional neural network, the fusion image is adopted to carry out graying and normalization operations, a gaussian filter is constructed to filter the fusion image, and data processing is carried out by a method of empirical mode decomposition and histogram equalization.
As a further alternative, the convolutional neural network comprises an input layer, a convolutional layer and a pooling layer, a batch normalization layer and an activation function layer, a fully connected layer design, and model optimization.
As a further alternative, the input layer is characterized by: assume that the size of the input fusion image is w×h, where W and H represent the width and height of the image, respectively;
convolution layer and pooling layer: extracting local features by using a smaller convolution kernel size, gradually extracting features of higher layers by using a plurality of convolution layers, and then performing dimension reduction processing on the features by using a pooling layer;
batch normalization layer and activation function layer: adding a batch normalization layer and an activation function layer between the convolution layer and the full connection layer, using a ReLU as an activation function, and adding the batch normalization layer to normalize the output;
full tie layer: using a full-connection layer with a plurality of neurons as an output layer and using a softmax function as an activation function to generate probability values of each category, representing the probability that the image belongs to different categories, defining 4 categories, and respectively corresponding to no-rust, slight-rust, moderate-rust and serious-rust, wherein the formula of the full-connection layer is as follows:
y = softmax(W*x + b)
wherein x represents an input feature, W and b represent weights and bias parameters of the full-connection layer, and y represents a probability value of each category;
model optimization: the cross entropy is used as a loss function, a random gradient descent algorithm is used for training and optimizing the model, and a regularization technology and a dropout technology are adopted to avoid overfitting, wherein the formula of the cross entropy is as follows:
Loss = -1/N * sum(Z*log(y_hat))
where N represents the number of samples, Z represents the true label, and y_hat represents the probability value predicted by the model.
In a third aspect, a stepper robot is provided for implementing the method described above.
In a fourth aspect, a computer storage medium has stored thereon a computer program which, when executed by a processor, implements the method described above.
Those skilled in the art will appreciate that all or part of the processes in the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, where the program may be stored in a non-volatile computer readable storage medium, and where the program, when executed, may include processes in the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink), DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
The foregoing has described in detail the embodiments of the present disclosure, so as to not obscure the technical idea of the present disclosure, and those skilled in the art will be able to implement the technical scheme of the disclosure based on the description of the embodiments.
The foregoing description of the preferred embodiments of the application is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the application.
Claims (8)
1. The device for detecting the rust of the chilled water pipe of the data center is characterized by comprising a stepping part, a marking part, a laser detection part, an infrared detection part, an environment reference part and an external processing host terminal;
the stepping part comprises a stepping device sleeved on a freezing water pipe heat preservation layer (2), the stepping device comprises a semi-annular body (3) and a component bracket (4), the semi-annular body comprises a fixed bridge (5) in the middle of the body and two extensible elastic ring arms (6), one ends of the fixed bridge, which deviate from the freezing water pipe heat preservation layer, are connected with the component bracket (4), the end parts of the two extensible elastic ring arms of the semi-annular body and the end, which is close to the freezing water pipe heat preservation layer, of the fixed bridge are respectively connected with an electric hydraulic connecting rod (7), the electric hydraulic connecting rods connected with the fixed bridge are provided with an active sliding wheel (8), a remotely adjustable stepping motor is arranged on the active sliding wheel, the electric hydraulic connecting rods corresponding to the end parts of the two extensible elastic ring arms are connected with driven wheels (9), and when the stepping motor is started, the active sliding wheel (8) drives the driven wheels (9) to cling to the freezing water pipe heat preservation layer to advance along the direction of the water pipe (1);
the marking part is arranged on at least one driven wheel (9), the marking part comprises a spraying device (10) and an encoder, the encoder is used for measuring relative displacement, and the spraying device (10) is used for spraying a plurality of positioning marks on the insulating layer (2) of the chilled water pipe at intervals according to preset interval displacement parameters;
the laser detection part (11) is positioned on the component bracket (4), and the laser detection part (11) comprises a laser emitter, a rotary scanner, a laser receiver, a clock and counter, a data processor and a first data wireless transmission module; the laser transmitter transmits laser beams, the rotary scanner scans the environment near the whole chilled water pipe in the horizontal and vertical directions, the laser receiver receives the laser beams and performs photoelectric conversion, the clock and the counter are used for measuring the round trip time of the laser beams and calculating the distance from an object in the environment to a laser radar, the data processor is used for processing the received laser signals of the object in the environment and the target laser signals and generating an accurate three-dimensional point cloud image, and the first data wireless transmission module transmits the three-dimensional point cloud image to the external processing host terminal;
the infrared detection part (12) is positioned on the component bracket (4) and comprises an infrared camera lens, an infrared sensing array, a data acquisition card and a second data wireless transmission module;
the environmental benchmark section comprises a target, which is a known benchmark pre-arranged in the vicinity of a data center chilled water pipe (1);
the external processing host terminal comprises a three-dimensional point cloud image characteristic processing unit, a thermal infrared characteristic processing unit, a fusion processing unit and a rust detection unit; the three-dimensional point cloud image feature processing unit sets a first feature recognition operator according to the positioning mark, the first feature recognition operator recognizes the positioning mark, and then determines the absolute coordinate of the positioning mark in the three-dimensional point cloud image by combining with the target; the thermal infrared characteristic processing unit sets a second characteristic recognition operator according to the positioning marks, the second characteristic recognition operator recognizes the positioning marks, when at least two positioning marks which are arranged at intervals are scanned and recognized on a thermal infrared image, the image pixel positions of the positioning marks in the thermal infrared image are determined, when at least one positioning mark is recognized by an adjacent thermal infrared image, the adjacent thermal infrared images are spliced to form a thermal infrared spliced image, and the fusion processing unit carries out fusion correction with the thermal infrared spliced image based on the positioning marks, the three-dimensional point cloud image and preset interval displacement parameters to form a fusion image; the rust detection unit constructs a convolutional neural network to process the fusion image obtained by the fusion processing unit so as to obtain a rust detection result;
wherein the first feature recognition operator recognizes the positioning mark, and combining the target, and then determining the absolute coordinates of the positioning mark in the three-dimensional point cloud image comprises: if more than two positioning marks are identified, vector connection is carried out between the positioning marks and the targets, adjustment is carried out by taking the preset interval displacement parameter as a distance constraint condition, an absolute coordinate correction value of the positioning marks in the three-dimensional point cloud picture is obtained according to the adjustment, and the absolute coordinates of the positioning marks in the three-dimensional point cloud picture are determined based on the absolute coordinate correction value.
2. The chilled water pipe rust detection device for data centers according to claim 1, wherein the infrared camera lens is arranged on one side of the component support close to the pipeline, the infrared sensing array receives heat radiation signals of the chilled water pipe acquired by the infrared camera lens, the data acquisition card is used for photoelectric conversion, and the second data wireless transmission module sends a plurality of heat infrared images acquired by the photoelectric conversion to the external processing host terminal.
3. The device for detecting the rust on a chilled water tube of a data center according to claim 1, wherein the positioning mark has a serial number coding function.
4. A rust detection method for detecting a rust of a chilled water tube of a data center using the device according to any one of claims 1 to 3, characterized in that:
step one: starting a stepping motor, wherein the driving sliding wheel (8) drives the driven wheel (9) to cling to the frozen water pipe heat insulation layer (2) to advance along the direction of the frozen water pipe; the infrared detection part (12) samples the frozen water pipe coated with the heat preservation layer according to a preset sampling frequency; the laser detection part (11) receives the laser signals of the environmental object and the target laser signals and generates an accurate three-dimensional point cloud picture; fusion correction is carried out on the thermal infrared spliced image based on the positioning mark, the three-dimensional point cloud image and the preset interval displacement parameter, so that a fusion image is formed;
secondly, carrying out graying and normalization operations on the fusion image, constructing a gaussian filter, filtering the fusion image, and carrying out data processing by a method of empirical mode decomposition combined with histogram equalization, wherein the specific data processing process is to decompose the fusion image into IMF1 to IMF6 through empirical mode decomposition, wherein the IMF1 has the highest frequency and the IMF6 has the lowest frequency, carrying out histogram equalization processing on IMF2-IMF6 components, setting a preset threshold value for IMF1 components, filtering high-frequency noise, and carrying out reconstruction and normalization on the IMF1-IMF6 to obtain a processed fusion image;
inputting the processed fusion image into a convolutional neural network for rust detection, classifying the rust degree according to rust characterization, wherein the convolutional neural network is established by acquiring a data center chilled water pipe infrared rust image labeling set as a test set, adopting the fusion image as a training set, training the convolutional neural network through the test set and the training set, and obtaining a rust detection result by utilizing the trained convolutional neural network.
5. The rust detection method according to claim 4, characterized in that: the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a batch normalization layer, an activation function layer, a full connection layer design and model optimization.
6. The rust detection method according to claim 5, characterized in that:
input layer: assume that the size of the input fusion image is w×h, where W and H represent the width and height of the image, respectively;
convolution layer and pooling layer: extracting local features by using a smaller convolution kernel size, gradually extracting features of higher layers by using a plurality of convolution layers, and then performing dimension reduction processing on the features by using a pooling layer;
batch normalization layer and activation function layer: adding a batch normalization layer and an activation function layer between the convolution layer and the full connection layer, using a ReLU as an activation function, and adding the batch normalization layer to normalize the output;
full tie layer: using a full-connection layer with a plurality of neurons as an output layer and using a softmax function as an activation function to generate probability values of each category, representing the probability that the image belongs to different categories, defining 4 categories, and respectively corresponding to no-rust, slight-rust, moderate-rust and serious-rust, wherein the formula of the full-connection layer is as follows:
y = softmax(W*x + b)
wherein x represents an input feature, W and b represent weights and bias parameters of the full-connection layer, and y represents a probability value of each category;
model optimization: the cross entropy is used as a loss function, a random gradient descent algorithm is used for training and optimizing the model, and a regularization technology and a dropout technology are adopted to avoid overfitting, wherein the formula of the cross entropy is as follows:
Loss = -1/N * sum(Z*log(y_hat))
where N represents the number of samples, Z represents the true label, and y_hat represents the probability value predicted by the model.
7. A stepper robot for implementing the method of any of claims 4-6.
8. A computer storage medium having stored thereon a computer program which, when executed by a processor, implements the method of any of claims 4-6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311047384.8A CN116754609B (en) | 2023-08-21 | 2023-08-21 | Method and device for detecting rust of chilled water pipe of data center |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311047384.8A CN116754609B (en) | 2023-08-21 | 2023-08-21 | Method and device for detecting rust of chilled water pipe of data center |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116754609A CN116754609A (en) | 2023-09-15 |
CN116754609B true CN116754609B (en) | 2023-12-05 |
Family
ID=87957614
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311047384.8A Active CN116754609B (en) | 2023-08-21 | 2023-08-21 | Method and device for detecting rust of chilled water pipe of data center |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116754609B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117744926B (en) * | 2023-12-20 | 2024-07-19 | 广东腾安机电安装工程有限公司 | Construction management method and system for assembled pipeline of refrigeration machine room |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008134221A (en) * | 2006-10-24 | 2008-06-12 | Nippon Steel Corp | Infrared pipe diagnostic method, and infrared pipe diagnostic device |
CN109218702A (en) * | 2018-09-05 | 2019-01-15 | 天目爱视(北京)科技有限公司 | A kind of camera rotation type 3D measurement and information acquisition device |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB201303324D0 (en) * | 2013-02-25 | 2013-04-10 | Subterandt Ltd | Passive detection of deformation under coatings |
-
2023
- 2023-08-21 CN CN202311047384.8A patent/CN116754609B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008134221A (en) * | 2006-10-24 | 2008-06-12 | Nippon Steel Corp | Infrared pipe diagnostic method, and infrared pipe diagnostic device |
CN109218702A (en) * | 2018-09-05 | 2019-01-15 | 天目爱视(北京)科技有限公司 | A kind of camera rotation type 3D measurement and information acquisition device |
Also Published As
Publication number | Publication date |
---|---|
CN116754609A (en) | 2023-09-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10768094B2 (en) | Thermography image processing with neural networks to identify corrosion under insulation (CUI) | |
US10871444B2 (en) | Inspection and failure detection of corrosion under fireproofing insulation using a hybrid sensory system | |
CN116754609B (en) | Method and device for detecting rust of chilled water pipe of data center | |
CN109685066B (en) | Mine target detection and identification method based on deep convolutional neural network | |
CN112348034A (en) | Crane defect detection system based on unmanned aerial vehicle image recognition and working method | |
CN113240887A (en) | Slope rockfall collapse monitoring method, device and equipment | |
WO2020123505A1 (en) | Inspection and failure detection of corrosion under fireproofing insulation using a hybrid sensory system | |
CN112697134A (en) | Environment sensing method, system and equipment of indoor inspection robot and computer readable storage medium thereof | |
CN112434569A (en) | Thermal imaging system of unmanned aerial vehicle | |
Gao et al. | Height-variable monocular vision ranging technology for smart agriculture | |
CN117704970A (en) | Building visual displacement monitoring system and monitoring method thereof | |
CN117647526A (en) | Wind driven generator blade surface defect detection method and device | |
CN111783726A (en) | Herding livestock outlier video acquisition method based on artificial intelligence and BIM | |
CN115294562B (en) | Intelligent sensing method for operation environment of plant protection robot | |
CN114982739A (en) | Intelligent laser bird repelling device and method based on deep learning | |
CN113191389B (en) | Submarine pipeline autonomous inspection method and device based on optical vision technology | |
CN109375212B (en) | Target detection method based on radar and photoelectric turntable in movable unmanned platform | |
CN117495938B (en) | Foldable hollow plate production data extraction method based on image processing | |
CN117809263B (en) | Intelligent block intelligent inspection control method, device and medium | |
CN117523532B (en) | Weed point cloud filtering method, system and storage medium | |
Rajavel et al. | Construction Progress Monitoring In Smart Cities Using Deep Alexnet | |
CN109034105B (en) | Tuna fishery position prediction method based on unmanned aerial vehicle | |
Jung et al. | Segmentation of Tomato Growing Truss a Depth Image Conversion Model Based on CycleGAN | |
CN118379640A (en) | Intelligent inspection method, system, equipment and medium for granary | |
CN114780771A (en) | Method and device for generating digital twin of environment, electronic device and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |