CN113536989B - Refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera video - Google Patents
Refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera video Download PDFInfo
- Publication number
- CN113536989B CN113536989B CN202110727931.1A CN202110727931A CN113536989B CN 113536989 B CN113536989 B CN 113536989B CN 202110727931 A CN202110727931 A CN 202110727931A CN 113536989 B CN113536989 B CN 113536989B
- Authority
- CN
- China
- Prior art keywords
- image
- model
- refrigerator
- condenser
- convolution
- 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
- 238000004458 analytical method Methods 0.000 title claims abstract description 32
- 238000012544 monitoring process Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 50
- 239000010410 layer Substances 0.000 claims description 36
- 230000006870 function Effects 0.000 claims description 28
- 238000004364 calculation method Methods 0.000 claims description 23
- 230000004913 activation Effects 0.000 claims description 19
- 238000011423 initialization method Methods 0.000 claims description 17
- 238000007781 pre-processing Methods 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 230000002457 bidirectional effect Effects 0.000 claims description 7
- 238000003709 image segmentation Methods 0.000 claims description 7
- 238000013136 deep learning model Methods 0.000 claims description 6
- 239000002346 layers by function Substances 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 4
- 230000008447 perception Effects 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims 4
- 230000007613 environmental effect Effects 0.000 description 13
- 235000013305 food Nutrition 0.000 description 10
- 238000005265 energy consumption Methods 0.000 description 8
- 238000003860 storage Methods 0.000 description 8
- 238000003745 diagnosis Methods 0.000 description 5
- 238000005057 refrigeration Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 230000007797 corrosion Effects 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 235000013622 meat product Nutrition 0.000 description 2
- 238000010223 real-time analysis Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000010257 thawing Methods 0.000 description 2
- 238000009827 uniform distribution Methods 0.000 description 2
- 235000013311 vegetables Nutrition 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 235000013365 dairy product Nutrition 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 235000013372 meat Nutrition 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000000275 quality assurance Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Image Analysis (AREA)
- Cold Air Circulating Systems And Constructional Details In Refrigerators (AREA)
Abstract
The invention provides a refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera videos, wherein the method comprises the following steps: step 1, obtaining video data of an area where a condenser in a refrigerator is located; step 2, separating a condenser area image from the video data; step 3, determining whether the separated condenser area image needs pretreatment, if so, entering a step 4 after pretreatment, otherwise, entering a step 6; step 4, judging whether the separated condenser region image needs image quality enhancement, if so, performing super-resolution image enhancement, and then entering a step 5, otherwise, entering a step 6; step 5, constructing a convolutional neural network model and predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model; and 6, constructing ConvLSTM a model and predicting the frosting condition of the refrigerator in real time by using the constructed ConvLSTM model. The invention can accurately diagnose the frosting condition of the refrigerator.
Description
Technical Field
The invention relates to the technical field of frosting diagnosis of refrigeration equipment, in particular to a refrigerator frosting monitoring method and system based on camera video frame-by-frame analysis.
Background
With the development of society and the improvement of living standard of residents, food quality assurance and food safety problems are increasingly receiving social importance, and food quality guarantee, storage and circulation are always one of the most important research contents in the fields of agriculture and food industry. In 2019, the import amount of frozen and refrigerated aquatic products and meat products in China rises to 1000 ten thousand tons, the total yield of fruits, vegetables, meat products, aquatic products and dairy products is expected to break through 13 hundred million tons, the market demand of a cold chain is huge, but the comprehensive cold chain circulation rate is only 17 percent (2.2563 hundred million tons), the corrosion rate is high (fruits, vegetables, meat and aquatic products respectively reach 20 percent to 30 percent, 12 percent and 15 percent), the direct economic loss caused by the direct economic loss exceeds 6800 hundred million yuan (accounting for 1 percent of GDP in the whole country) and the great waste of social resources is caused. The cold chain logistics can provide a proper temperature environment for the storage and circulation of perishable foods, and is a key for reducing the corrosion rate, maintaining the food quality and safety.
The refrigeration house is an important infrastructure for food freezing processing, storage and circulation, is a key node of a whole-course cold chain, has irreplaceable functions in the aspects of cold chain commodity storage and quality guarantee, plays an important role in national economy, and is huge in total quantity. The total amount of the cold storage in 2018 nationwide reaches 5238 ten thousand tons (about 1.3 hundred million cubic meters), the total amount of logistics exceeds 4 trillion yuan, and the construction and technical development of the cold storage are important research contents in the fields of food industry and logistics management. However, with the rapid development of cold-chain logistics represented by cold storage, the energy consumption of the cold-chain logistics is rapidly increased, and taking the food industry as an example, the energy consumption of a refrigeration system comprising the steps of production, circulation and storage is 35.0% of the total energy consumption of the food industry, and the total energy consumption reaches 1300 TWh/year in the global scope, so that the energy consumption of the cold-chain logistics is a serious challenge. And because the environment in the logistics process needs to be controlled, the cost of the cold chain logistics is 40.0% higher than that of the common logistics, and the problems of high cost and low efficiency are particularly remarkable. In each link of the cold chain logistics, the working efficiency of the refrigerating system is a key factor for increasing the refrigerating effect and reducing the energy consumption, and the refrigerating efficiency is severely restricted because the air outlet of the refrigerating machine is extremely easy to frost in a low-temperature environment, so that the refrigerating of products of the logistics system can be influenced, and the energy consumption cost is greatly increased.
Therefore, the intelligent refrigerator working state diagnosis method is designed, and has important significance for improving the refrigeration efficiency and reducing the energy consumption cost. However, the traditional sensor monitoring mode can only reflect the temperature and the humidity of the environment, and can not accurately diagnose the frosting condition of the refrigerator.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a refrigerator frosting monitoring method based on the frame-by-frame analysis of a shooting video, which can accurately diagnose the frosting condition of a refrigerator.
In order to achieve the above purpose, the invention provides a refrigerator frosting monitoring method based on the frame-by-frame analysis of a camera video, which comprises the following steps:
step 1, obtaining video data of an area where a condenser in a refrigerator is located;
Step 2, separating a condenser area image from the video data according to the video data;
Step 3, determining whether the separated condenser area image needs pretreatment, if so, entering a step 4 after pretreatment, otherwise, entering a step 6;
Step 4, judging whether the separated condenser region image needs image quality enhancement, if so, performing super-resolution image enhancement on the separated condenser region image, and then entering a step 5, otherwise, entering a step 6;
step 5, constructing a convolutional neural network model according to the processed condenser region image data, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model;
And 6, constructing ConvLSTM a model according to the separated condenser area image, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM model.
Further, in step 2, the step of separating the condenser region image from the video data includes:
step 201, constructing a partitioned convolutional neural network model operation environment;
Step 202, performing convolution calculation on an input image to realize downsampling of the image, performing multi-scale feature fusion calculation, and inputting the abstract features obtained by calculation to a next convolution layer;
And 203, performing transpose convolution calculation on the input of the next convolution layer to realize up-sampling of the abstract feature, restoring the size of the abstract feature to the original image size, and reserving a data area of the abstract feature, so as to divide the condenser area image.
Further, in step 3, the step of determining whether the separated condenser area image needs preprocessing includes:
Step 301, calculating noise points existing in the image, wherein the noise points comprise but are not limited to noise points formed by Gaussian noise and impulse noise;
Step 302, counting the noise image area according to the noise points, and judging whether the noise image area is within a preset threshold range or not; if yes, pretreatment is needed, otherwise, pretreatment is not needed.
Further, in step 4, the step of determining whether the separated condenser region image requires image enhancement includes: and counting the average pixel value of the separated condenser region image, judging whether the average pixel value is within a preset pixel average value range, if so, carrying out image enhancement, otherwise, not carrying out image enhancement.
Further, in step 4, the step of performing super-resolution image enhancement on the separated condenser region image includes:
Step 401, constructing a partitioned convolutional neural network model operating environment;
Step 402, extracting target image features by setting a convolution layer in a deep learning model, and constructing a target feature set;
Step 403, calculating the distance between the target feature set and the generated feature set in the high-dimensional feature space through the perception loss function, and calculating the optimal optimization gradient by using a back propagation algorithm to optimize the target image features;
step 404, upsampling the optimized features to output a resolution enhanced image.
Further, in step 5, the step of constructing a convolutional neural network model according to the processed condenser region image data includes:
Step 501, constructing a convolutional neural network model with a sequential structure, wherein the convolutional neural network model comprises 2 groups of convolutional modules, and each convolutional module comprises 2 convolutional layers and 1 pooled sampling layer;
step 502, selecting optimal items of a model initialization method, an activation function and an optimizer;
in step 503, the optimal convolution kernel is calculated and determined.
Further, in step 6, the step of constructing ConvLSTM a model according to the separated condenser region image includes:
step 601, constructing ConvLSTM models of sequential structures, and setting ConvLSTM functional layers in the models;
step 602, calculating convolution operation of the image under each time sequence by using TimeDistributed packagers;
Step 603, calculating the logic relation of the image in the time sequence direction by using a bidirectory Bidirectional wrapper, and extracting the time sequence characteristics of the image data;
step 604, selecting optimal items of a model initialization method, an activation function and an optimizer;
In step 605, other model parameters are computationally determined.
On the other hand, the invention also provides a refrigerator frosting monitoring system based on the frame-by-frame analysis of the camera video, which comprises
The video data acquisition module is used for acquiring video data of the area where the condenser in the refrigerator is located;
The image segmentation module is used for separating a condenser area image from the video data according to the video data;
The image preprocessing module is used for determining whether the separated condenser area image needs preprocessing or not, and if so, preprocessing is carried out;
the image quality enhancement module is used for judging whether the separated condenser region image needs image quality enhancement or not, if so, super-resolution image enhancement is carried out on the separated condenser region image;
The first prediction module is used for constructing a convolutional neural network model for the condenser region image data subjected to super-resolution image enhancement by the image quality enhancement module, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model;
And the second prediction module is used for constructing ConvLSTM a model for the condenser area image which does not need to be preprocessed and/or subjected to the image quality enhancement module, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM model.
Further, the first prediction module comprises
The convolution neural network model building unit is used for building convolution neural network modules with sequential structures, wherein the convolution neural network model comprises 2 groups of convolution modules, and each convolution module comprises 2 convolution layers and 1 pooled sampling layer;
the first optimal item selection unit is used for selecting optimal items of a model initialization method, an activation function and an optimizer;
And the convolution kernel calculation unit is used for calculating and determining the optimal convolution kernel.
Further, the second prediction module includes
ConvLSTM a model construction unit, which is used for constructing a ConvLSTM model of the sequential structure and arranging ConvLSTM functional layers in the model;
A convolution operation calculation unit for calculating a convolution operation of the image at each time sequence by TimeDistributed wrappers;
The time sequence feature determining unit is used for calculating the logic relation of the image in the time sequence direction by using the bidirectory Bidirectional encapsulator and extracting the time sequence feature of the image data;
the second optimal item selection unit is used for selecting optimal items of a model initialization method, an activation function and an optimizer;
And the model parameter calculation unit is used for calculating and determining other model parameters.
Compared with the prior art, the invention has the following advantages: the invention collects the image of the air outlet condenser area by adopting a camera video monitoring mode, and carries out frosting area segmentation and image super-resolution enhancement on the image quality by adopting a video image frame-by-frame analysis method, thereby realizing an accurate judging method for the frosting condition of the air outlet condenser of the refrigerating machine system. Specifically, when environmental interference exists in the image, the frosting condition of the refrigerator is predicted by constructing a convolutional neural network model; when the image is not interfered, the frost forming condition of the refrigerator is predicted by constructing ConvLSTM models, the defects that monitoring equipment is easily affected by the environment of a refrigerator, the definition of image quality is insufficient, the video information is difficult to extract and the like are overcome, and the purpose of real-time and accurate monitoring of the working state of the refrigerator is realized at low cost.
In addition, the convolutional neural network group model based on deep learning can effectively process video time sequence data, and can realize accurate and real-time frosting diagnosis analysis by combining video image content change analysis and real-time analysis information in a small range of time, thereby providing timely and accurate intelligent early warning functions for defrosting and manual intervention of an air outlet condenser.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a flow chart of a refrigerator frosting monitoring method based on a camera video frame-by-frame analysis of the invention;
Fig. 2 is a block diagram of a refrigerator frosting monitoring system based on frame-by-frame analysis of a camera video.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the embodiment of the invention discloses a refrigerator frosting monitoring method based on shooting video frame-by-frame analysis, which comprises the following steps:
step 1, obtaining video data of an area where a condenser in a refrigerator is located;
Step 2, separating a condenser area image from the video data according to the video data;
Step 3, determining whether the separated condenser area image needs pretreatment, if so, entering a step 4 after pretreatment, otherwise, entering a step 6;
Step 4, judging whether the separated condenser region image needs image quality enhancement, if so, performing super-resolution image enhancement on the separated condenser region image, and then entering a step 5, otherwise, entering a step 6;
step 5, constructing a convolutional neural network model according to the processed condenser region image data, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model;
And 6, constructing ConvLSTM a model according to the separated condenser area image, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM model.
Correspondingly, referring to fig. 2, the embodiment of the invention also discloses a refrigerator frosting monitoring system based on the image pickup video frame-by-frame analysis, which comprises
The video data acquisition module is used for acquiring video data of the area where the condenser in the refrigerator is located;
The image segmentation module is used for separating a condenser area image from the video data according to the video data;
The image preprocessing module is used for determining whether the separated condenser area image needs preprocessing or not, and if so, preprocessing is carried out;
the image quality enhancement module is used for judging whether the separated condenser region image needs image quality enhancement or not, if so, super-resolution image enhancement is carried out on the separated condenser region image;
The first prediction module is used for constructing a convolutional neural network model for the condenser region image data subjected to super-resolution image enhancement by the image quality enhancement module, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model;
And the second prediction module is used for constructing ConvLSTM a model for the condenser area image which does not need to be preprocessed and/or subjected to the image quality enhancement module, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM model.
In this embodiment, the refrigerator frost monitoring method based on the captured video frame-by-frame analysis uses the refrigerator frost monitoring system based on the captured video frame-by-frame analysis as an execution object of the step, or uses the component parts in the refrigerator frost monitoring system based on the captured video frame-by-frame analysis as an execution object of the step. Specifically, step 1 uses a video data acquisition module as an execution object of the step, step 2 uses an image segmentation module as an execution object of the step, step 3 uses an image preprocessing module as an execution object of the step, step 4 uses an image quality enhancement module as an execution object of the step, step 5 uses a first prediction module as an execution object of the step, and step 6 uses a second prediction module as an execution object of the step.
Because the frosting of the condenser (or the evaporator) can block the air duct and obviously affect the working efficiency of the refrigerator, the invention uses the camera to monitor the frosting condition of the condenser area, thereby diagnosing the working efficiency of the refrigerator. However, the image obtained by using the camera is easily interfered by environmental factors in the low-temperature environment of the refrigerator, and the problems of low resolution, poor imaging effect, incapability of intelligent identification and the like occur; aiming at the situation, the method analyzes based on video data acquired by a camera, judges whether the image acquired by the camera in real time has environmental interference, if so, performs super-resolution image enhancement on the separated condenser area image, constructs a convolutional neural network model by utilizing the condenser area image data, and predicts the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model; if the image acquired by the camera in real time has no environmental interference, constructing ConvLSTM models according to the separated condenser area images, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM models; according to the invention, different deep learning models are adopted to predict the frosting condition of the refrigerator according to whether the environment interference exists, so that the frosting condition of the refrigerator can be accurately diagnosed.
In step 1, video data of an area where a condenser in a refrigerator is located, which is acquired by a camera, is acquired.
In step 2, the video frame collected by the camera is segmented to separate out the condenser area image, so that other irrelevant factors in the image are prevented from interfering with the subsequent analysis.
Specifically, in step 2, the step of separating the condenser region image from the video data includes:
step 201, constructing a partitioned convolutional neural network model operation environment;
Step 202, performing convolution calculation on an input image to realize downsampling of the image, performing multi-scale feature fusion calculation through feature point-by-point addition and feature channel dimension splicing, and then inputting the calculated abstract features to a next convolution layer;
And 203, performing transpose convolution calculation on the input of the next convolution layer to realize up-sampling of the abstract feature, restoring the size of the abstract feature to the original image size, and reserving a data area of the abstract feature, so as to divide the condenser area image.
In the embodiment of the invention, a segmentation convolutional neural network model is constructed to downsample and upsample the image, the acquired multi-scale features are fused through feature point-by-point addition and feature channel dimension splicing, and each pixel point is judged in category, so that the image segmentation result at the pixel level is realized.
Because the network can accept pictures with any size and output the segmentation pictures with the same size or larger than the original pictures, a transposition convolution layer is arranged in the network structure, the feature pictures are mapped back to the original pictures with the size or larger, and the large feature pictures are rolled out from the small feature pictures.
The conventional convolutional neural network uses a rounding approximation to find the corresponding model parameters of the image region in the network structure when processing image segmentation, which can cause the offset of the corresponding relation from the actual situation. This offset is typically negligible in image classification tasks, but may have a larger impact in higher definition image segmentation tasks. Therefore, in this embodiment, by using the region feature aggregation manner, the step 202 is to perform convolution calculation on any region in the input image corresponding to the corresponding region in the neural network feature map to implement downsampling of the image, so as to aggregate the region features, thereby solving the problem of region mismatch caused by two times of quantization in the conventional convolutional neural network.
In step 3, after the segmented condenser area image is obtained, whether the segmented condenser area image needs to be preprocessed or not needs to be determined, namely whether noise exists or not, and if not, the frost situation can be predicted by constructing ConvLSTM models; if noise exists, a further determination is required.
Specifically, in step 3, the step of determining whether the separated condenser area image needs preprocessing includes:
Step 301, calculating noise points existing in the image, wherein the noise points comprise but are not limited to noise points formed by Gaussian noise and impulse noise;
Step 302, counting the noise image area according to the noise points, and judging whether the noise image area is within a preset threshold range or not; if yes, pretreatment is needed, otherwise, pretreatment is not needed.
In the embodiment of the invention, after the segmented condenser region image is obtained, noise points existing in the image are calculated, the noise image area is counted, and when the noise image area is larger than a preset threshold range, the acquired image is possibly interfered by environment and needs further judgment; if the area of the noise image is within the preset threshold, the collected image is proved to have no environmental interference, so that the frost situation can be predicted directly by constructing ConvLSTM models.
In step 4, after removing noise from the separated condenser area image, it is further required to determine whether the image has environmental interference, specifically, whether the separated condenser area image needs image enhancement is determined to have environmental interference.
Specifically, in step 4, the step of determining whether the separated condenser region image requires image enhancement includes: and counting the average pixel value of the separated condenser region image, judging whether the average pixel value is within a preset pixel average value range, if so, carrying out image enhancement, otherwise, not carrying out image enhancement.
In the embodiment of the invention, the determination of the preset pixel average value range is performed by calculating the pixel average value of a plurality of images with environmental interference; if the average pixel value of the separated condenser region image is within the preset pixel average value range, the condition that the image has environmental interference is proved, so that the super-resolution image enhancement is needed to be carried out on the image, and then a convolutional neural network model is built for predicting frosting condition.
Specifically, in step 4, the step of performing super-resolution image enhancement on the separated condenser region image includes:
Step 401, constructing a partitioned convolutional neural network model operating environment;
Step 402, extracting target image features by setting a convolution layer in a deep learning model, and constructing a target feature set;
Step 403, calculating the distance between the target feature set and the generated feature set in the high-dimensional feature space through the perception loss function, and calculating the optimal optimization gradient by using a back propagation algorithm to optimize the target image features;
step 404, upsampling the optimized features to output a resolution enhanced image.
Since conventional convolutional neural networks typically use the mean square error as a loss function in training the network, while guaranteeing a high peak signal-to-noise ratio, the resulting image typically loses high frequency detail and cannot be used to identify the degree of frosting of the condenser (or evaporator). In the embodiment of the invention, in order to promote the feature details of the restored picture, the difference between the current model weight and the target expectation is calculated by adopting the perception loss, and the training effect is achieved by continuously correcting the model coefficient; specifically, the characteristics of the condenser (or evaporator) region image extracted by the convolution layer are compared with the differences of the characteristics of the generated picture after passing through the convolution neural network and the characteristics of the target picture after passing through the convolution neural network, so that the generated picture and the target picture are more similar in terms of semantics and style.
In step 402, a convolution layer is required to be provided in the deep learning model, and a plurality of convolution layers are provided in the convolution module in consideration of the pixel size of the evaporator (or condenser) region, and the convolution kernels are respectively set to 3×3,5×5, and 7×7 in size.
In step 5, if the acquired image has environmental interference, constructing a convolutional neural network model to predict the frosting condition of the refrigerator after the super-resolution enhancement of the condenser area image with the environmental interference.
Specifically, in step 5, the step of constructing a convolutional neural network model according to the processed condenser region image data includes:
Step 501, constructing a convolutional neural network model with a sequential structure, wherein the convolutional neural network model comprises 2 groups of convolutional modules, and each convolutional module comprises 2 convolutional layers and 1 pooled sampling layer;
step 502, selecting optimal items of a model initialization method, an activation function and an optimizer;
in step 503, the optimal convolution kernel is calculated and determined.
Correspondingly, in the refrigerator frosting monitoring system based on the frame-by-frame analysis of the camera video, the first prediction module comprises
The convolution neural network model building unit is used for building convolution neural network modules with sequential structures, wherein the convolution neural network model comprises 2 groups of convolution modules, and each convolution module comprises 2 convolution layers and 1 pooled sampling layer;
the first optimal item selection unit is used for selecting optimal items of a model initialization method, an activation function and an optimizer;
And the convolution kernel calculation unit is used for calculating and determining the optimal convolution kernel.
In this embodiment, step 5 takes the first prediction module as an execution object of the step, or takes the component part of the first prediction module as an execution object of the step. Specifically, step 501 takes a convolutional neural network model building unit as an execution object of the step, step 502 takes a first optimal item selecting unit as an execution object of the step, and step 503 takes a convolutional kernel number calculating unit as an execution object of the step.
In the embodiment of the invention, considering that many operations in the CNN model in the embodiment are based on a 2/4 grouping structure, the 2/4 performance is superior and stable in large-size image analysis, so that when the convolutional neural network model with a sequential structure is used, the convolutional neural network model comprises 2 groups of convolutional modules, and each convolutional module comprises 2 convolutional layers and 1 pooled sampling layer, so that the stability of the prediction model is improved.
And constructing a convolutional neural network model, determining an initialization method, an activation function and an optimizer of the model, and calculating the optimal convolution kernel number by using the determined initialization method, activation function and optimizer so as to finally determine the convolutional neural network model.
Specifically, the initialization method comprises uniform distribution initialization, all-0 initialization, all-1 initialization, fixed value initialization, normal distribution initialization, random uniform distribution initialization, truncated Gaussian distribution initialization, random orthogonal matrix initialization, unit matrix initialization and the like, and by using 2000 training image data, the model adopts each optimizer to train 2000 rounds in an iterative manner so as to study the influence of different initialization methods on the performance of the model; and determining an optimal initialization method by comprehensively considering the loss function and the accuracy of the test set.
For the selection of the activation function, the embodiment of the invention compares softmax, relu, sigmoid and LeakyReLu functions commonly used in the deep learning model as the activation functions to evaluate the best activation function performance. The optimal activation function is determined by comparing whether there is a non-0 output when the activation function is not activated.
For the selection of the optimizer, the embodiment of the invention respectively analyzes the performance of the optimizer for seven algorithms, namely RMSprop algorithm, adam algorithm, random gradient descent algorithm, adagrad algorithm, adadelta algorithm, adamax algorithm and Nadam algorithm.
In step 6, if the acquired image has no environmental interference, a ConvLSTM model is constructed to predict the frosting condition of the refrigerator.
Specifically, in step 6, the step of constructing ConvLSTM a model according to the separated condenser region image includes:
step 601, constructing ConvLSTM models of sequential structures, and setting ConvLSTM functional layers in the models;
step 602, calculating convolution operation of the image under each time sequence by using TimeDistributed packagers;
Step 603, calculating the logic relation of the image in the time sequence direction by using a bidirectory Bidirectional wrapper, and extracting the time sequence characteristics of the image data;
step 604, selecting optimal items of a model initialization method, an activation function and an optimizer;
In step 605, other model parameters are computationally determined.
Correspondingly, in the refrigerator frosting monitoring system based on the frame-by-frame analysis of the camera video, the second prediction module comprises
ConvLSTM a model construction unit, which is used for constructing a ConvLSTM model of the sequential structure and arranging ConvLSTM functional layers in the model;
A convolution operation calculation unit for calculating a convolution operation of the image at each time sequence by TimeDistributed wrappers;
The time sequence feature determining unit is used for calculating the logic relation of the image in the time sequence direction by using the bidirectory Bidirectional encapsulator and extracting the time sequence feature of the image data;
the second optimal item selection unit is used for selecting optimal items of a model initialization method, an activation function and an optimizer;
And the model parameter calculation unit is used for calculating and determining other model parameters.
In this embodiment, step 6 uses the second prediction module as an execution object of the step, or uses a component of the second prediction module as an execution object of the step. Specifically, step 601 uses ConvLSTM model building units as the execution object of the steps, step 602 uses convolution operation calculation units as the execution object of the steps, step 603 uses timing characteristic determination units as the execution object of the steps, step 604 uses second optimal item selection units as the execution object of the steps, and step 605 uses model parameter calculation units as the execution object of the steps.
In the embodiment of the invention, in order to increase the learning effect of the model on the time dimension, the feedback and the transverse connection from top to bottom are added to the feedforward connection from bottom to top, so that the visual time representation with strong robustness is provided; the LSTM layer is used for analyzing the time sequence image content association (time characteristic), the convolution layer is used for extracting the image content characteristics (space characteristic), the keras deep learning framework is used for combining the LSTM layer and the CNN layer to form ConvLSTM layers, and the time-space characteristics can be utilized simultaneously. The ConvLSTM core is essentially the same as LSTM, with the output of the upper layer being the input of the lower layer, but its input transform and cyclic transform are implemented by convolution. The difference from CNN is that not only can the timing relationship be obtained after LSTM plus convolution operation, but also features can be extracted like a convolution layer, spatial features can be extracted, and state-to-state switching is also changed to convolution computation. In ConvLSTM model, a layer is also required to be applied to each time slice of the input, and each time sequence of the time dimension is independently convolved to extract the characteristics, which is realized by a TimeDistributed wrapper provided by keras; the unidirectional LSTM is expanded, learning parameters are added during forward propagation, information extraction characteristics of a later sequence (future) are utilized, and the unidirectional LSTM is realized by using a Bidirectional encapsulator.
In the embodiment of the invention, for the collected monitoring video, on one hand, the brightness and the chromaticity values of pixels in two adjacent frames of images are relatively close, and as frosting is an image gradual change process which is as long as tens of seconds or even hundreds of seconds in the video, the situation that the short time mutation is larger in picture content difference can not occur, so early warning of early frosting can be carried out through a frame memory prediction method, namely a convolutional neural network model, at the early stage of frosting; on the other hand, when the environment suddenly interferes, the influence of the environment interference factor can be primarily eliminated through the image super-resolution processing, and for the interference which cannot be eliminated, the frame memory prediction technology is required, namely, the ConvLSTM model is used for diagnosing the frosting condition under the interference condition. Because the deep learning network lacks time variable existing in the video stream, smooth transition of scenes in the video cannot be realized; thus, a feed-forward deep neural network that uses data and tags from a large number of static images for supervised training is not suitable for frost diagnosis of condenser areas in refrigeration units.
In summary, the method for judging the frosting condition of the condenser at the air outlet of the refrigerator system is realized by adopting a camera video monitoring mode to collect the image of the condenser area at the air outlet and adopting a video image frame-by-frame analysis method to segment the frosting area and enhance the super-resolution of the image for the image quality. Specifically, when environmental interference exists in the image, the frosting condition of the refrigerator is predicted by constructing a convolutional neural network model; when the image is not interfered, the frost forming condition of the refrigerator is predicted by constructing ConvLSTM models, the defects that monitoring equipment is easily affected by the environment of a refrigerator, the definition of image quality is insufficient, the video information is difficult to extract and the like are overcome, and the purpose of real-time and accurate monitoring of the working state of the refrigerator is realized at low cost.
In addition, the convolutional neural network group model based on deep learning can effectively process video time sequence data, and can realize accurate and real-time frosting diagnosis analysis by combining video image content change analysis and real-time analysis information in a small range of time, thereby providing timely and accurate intelligent early warning functions for defrosting and manual intervention of an air outlet condenser.
The foregoing description of the preferred embodiments of the invention 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 invention.
Claims (9)
1. The refrigerator frosting monitoring method based on the frame-by-frame analysis of the camera video is characterized by comprising the following steps of:
step 1, obtaining video data of an area where a condenser in a refrigerator is located;
Step 2, separating a condenser area image from the video data according to the video data;
Step 3, determining whether the separated condenser area image needs pretreatment, if so, entering a step 4 after pretreatment, otherwise, entering a step 6;
Step 4, judging whether the separated condenser region image needs image quality enhancement, if so, performing super-resolution image enhancement on the separated condenser region image, and then entering a step 5, otherwise, entering a step 6;
step 5, constructing a convolutional neural network model according to the processed condenser region image data, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model;
Step 6, constructing ConvLSTM a model according to the separated condenser area image, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM model;
wherein in step 3, the step of determining whether the separated condenser area image requires preprocessing comprises:
Step 301, calculating noise points existing in the image, wherein the noise points comprise but are not limited to noise points formed by Gaussian noise and impulse noise;
Step 302, counting the noise image area according to the noise points, and judging whether the noise image area is within a preset threshold range or not; if yes, pretreatment is needed, otherwise, pretreatment is not needed.
2. The chiller frost monitoring method based on a frame-by-frame analysis of a captured video of claim 1, wherein in step 2, the step of separating the condenser area image from the video data comprises:
step 201, constructing a partitioned convolutional neural network model operation environment;
Step 202, performing convolution calculation on an input image to realize downsampling of the image, performing multi-scale feature fusion calculation, and inputting the abstract features obtained by calculation to a next convolution layer;
And 203, performing transpose convolution calculation on the input of the next convolution layer to realize up-sampling of the abstract feature, restoring the size of the abstract feature to the original image size, and reserving a data area of the abstract feature, so as to divide the condenser area image.
3. The method for monitoring frost formation on a refrigerator based on a frame-by-frame analysis of a captured video according to claim 1, wherein in step 4, the step of determining whether the separated condenser area image requires image enhancement comprises: and counting the average pixel value of the separated condenser region image, judging whether the average pixel value is within a preset pixel average value range, if so, carrying out image enhancement, otherwise, not carrying out image enhancement.
4. The method for monitoring frost formation on a refrigerator based on a frame-by-frame analysis of a captured video according to claim 1, wherein in step 4, the step of super-resolution image enhancement of the separated condenser area image includes:
Step 401, constructing a partitioned convolutional neural network model operating environment;
Step 402, extracting target image features by setting a convolution layer in a deep learning model, and constructing a target feature set;
Step 403, calculating the distance between the target feature set and the generated feature set in the high-dimensional feature space through the perception loss function, and calculating the optimal optimization gradient by using a back propagation algorithm to optimize the target image features;
step 404, upsampling the optimized features to output a resolution enhanced image.
5. The method for monitoring frost formation on a refrigerator based on a frame-by-frame analysis of a captured video according to claim 1, wherein in step 5, the step of constructing a convolutional neural network model from the processed condenser region image data comprises:
Step 501, constructing a convolutional neural network model with a sequential structure, wherein the convolutional neural network model comprises 2 groups of convolutional modules, and each convolutional module comprises 2 convolutional layers and 1 pooled sampling layer;
step 502, selecting optimal items of a model initialization method, an activation function and an optimizer;
in step 503, the optimal convolution kernel is calculated and determined.
6. The method for monitoring frost formation on a refrigerator based on a frame-by-frame analysis of a captured video according to claim 1, wherein in step 6, the step of constructing ConvLSTM model from the separated condenser area image comprises:
step 601, constructing ConvLSTM models of sequential structures, and setting ConvLSTM functional layers in the models;
step 602, calculating convolution operation of the image under each time sequence by using TimeDistributed packagers;
Step 603, calculating the logic relation of the image in the time sequence direction by using a bidirectory Bidirectional wrapper, and extracting the time sequence characteristics of the image data;
step 604, selecting optimal items of a model initialization method, an activation function and an optimizer;
In step 605, other model parameters are computationally determined.
7. Refrigerator frosting monitored control system based on make a video recording video frame by frame analysis, characterized by comprising
The video data acquisition module is used for acquiring video data of the area where the condenser in the refrigerator is located;
The image segmentation module is used for separating a condenser area image from the video data according to the video data;
the image preprocessing module is used for calculating noise points in the image, wherein the noise points comprise but are not limited to noise points formed by Gaussian noise and impulse noise, counting the noise image area according to the noise points, and judging whether the noise image area is in a preset threshold range or not; if yes, pretreatment is needed, otherwise, pretreatment is not needed;
the image quality enhancement module is used for judging whether the separated condenser region image needs image quality enhancement or not, if so, super-resolution image enhancement is carried out on the separated condenser region image;
The first prediction module is used for constructing a convolutional neural network model for the condenser region image data subjected to super-resolution image enhancement by the image quality enhancement module, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed convolutional neural network model;
And the second prediction module is used for constructing ConvLSTM a model for the condenser area image which does not need to be preprocessed and/or subjected to the image quality enhancement module, and then predicting the frosting condition of the refrigerator in real time by utilizing the constructed ConvLSTM model.
8. The chiller frost monitoring system based on a frame-by-frame analysis of a captured video of claim 7, wherein the first prediction module comprises
The convolution neural network model building unit is used for building convolution neural network modules with sequential structures, wherein the convolution neural network model comprises 2 groups of convolution modules, and each convolution module comprises 2 convolution layers and 1 pooled sampling layer;
the first optimal item selection unit is used for selecting optimal items of a model initialization method, an activation function and an optimizer;
And the convolution kernel calculation unit is used for calculating and determining the optimal convolution kernel.
9. The chiller frost monitoring system based on a frame-by-frame analysis of a captured video of claim 7, wherein the second prediction module comprises
ConvLSTM a model construction unit, which is used for constructing a ConvLSTM model of the sequential structure and arranging ConvLSTM functional layers in the model;
A convolution operation calculation unit for calculating a convolution operation of the image at each time sequence by TimeDistributed wrappers;
The time sequence feature determining unit is used for calculating the logic relation of the image in the time sequence direction by using the bidirectory Bidirectional encapsulator and extracting the time sequence feature of the image data;
the second optimal item selection unit is used for selecting optimal items of a model initialization method, an activation function and an optimizer;
And the model parameter calculation unit is used for calculating and determining other model parameters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110727931.1A CN113536989B (en) | 2021-06-29 | 2021-06-29 | Refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110727931.1A CN113536989B (en) | 2021-06-29 | 2021-06-29 | Refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113536989A CN113536989A (en) | 2021-10-22 |
| CN113536989B true CN113536989B (en) | 2024-06-18 |
Family
ID=78097155
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110727931.1A Active CN113536989B (en) | 2021-06-29 | 2021-06-29 | Refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113536989B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116989510A (en) * | 2023-09-28 | 2023-11-03 | 广州冰泉制冷设备有限责任公司 | Intelligent refrigeration method combining frosting detection and hot gas defrosting |
Family Cites Families (49)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101860677A (en) * | 2010-06-04 | 2010-10-13 | 西安天和防务技术股份有限公司 | Fog-penetrating and night-vision system based on refrigeration type CCD element |
| CN105654522B (en) * | 2015-12-30 | 2018-09-04 | 青岛海尔股份有限公司 | The frosting detection method and frosting detecting system of evaporator |
| US11545013B2 (en) * | 2016-10-26 | 2023-01-03 | A9.Com, Inc. | Customizable intrusion zones for audio/video recording and communication devices |
| WO2018170421A1 (en) * | 2017-03-17 | 2018-09-20 | Magic Leap, Inc. | Room layout estimation methods and techniques |
| CN107091548A (en) * | 2017-05-23 | 2017-08-25 | 天津大学 | A kind of air source heat pump defrosting control system and method |
| US11531994B2 (en) * | 2017-07-21 | 2022-12-20 | Vispera Bilgi Teknolojileri Sanayi Ic Ve Disticaret A.S. | Electronic detection of products and arrangement of products in a display structure, electronic detection of objects and arrangement of objects on and around the display structure, electronic detection of conditions of and around the display structure, and electronic scoring of the detected product and object arrangements and of the detected conditions |
| US11257259B2 (en) * | 2017-08-15 | 2022-02-22 | Siemens Healthcare Gmbh | Topogram prediction from surface data in medical imaging |
| CN107860099B (en) * | 2017-09-15 | 2018-11-30 | 珠海格力电器股份有限公司 | A kind of frosting detection method, device, storage medium and the equipment of air-conditioning |
| CN108009493B (en) * | 2017-11-30 | 2021-07-06 | 电子科技大学 | Human face anti-cheating recognition method based on motion enhancement |
| CN108038846A (en) * | 2017-12-04 | 2018-05-15 | 国网山东省电力公司电力科学研究院 | Transmission line equipment image defect detection method and system based on multilayer convolutional neural networks |
| US11534136B2 (en) * | 2018-02-26 | 2022-12-27 | Siemens Medical Solutions Usa, Inc. | Three-dimensional segmentation from two-dimensional intracardiac echocardiography imaging |
| US11537139B2 (en) * | 2018-03-15 | 2022-12-27 | Nvidia Corporation | Determining drivable free-space for autonomous vehicles |
| CN108844594B (en) * | 2018-04-17 | 2021-03-12 | 广州博通信息技术有限公司 | Alarm method and system for liquid level identification of air conditioner compressor |
| CN109308519A (en) * | 2018-09-29 | 2019-02-05 | 广州博通信息技术有限公司 | A kind of refrigeration equipment failure prediction method neural network based |
| CN109801294A (en) * | 2018-12-14 | 2019-05-24 | 深圳先进技术研究院 | Three-dimensional atrium sinistrum dividing method, device, terminal device and storage medium |
| CN109612027B (en) * | 2018-12-18 | 2023-10-27 | 天津大学 | Method and control system for analyzing frosting of air source heat pump based on microscopic photographing |
| CN109886105B (en) * | 2019-01-15 | 2021-12-14 | 广州图匠数据科技有限公司 | Price tag identification method, system and storage medium based on multi-task learning |
| CN110175538A (en) * | 2019-05-10 | 2019-08-27 | 国网福建省电力有限公司龙岩供电公司 | A kind of substation's Bird's Nest recognition methods and system based on machine learning |
| CN110363327B (en) * | 2019-06-04 | 2021-09-03 | 东南大学 | ConvLSTM and 3D-CNN-based short rainfall prediction method |
| CN110210429B (en) * | 2019-06-06 | 2022-11-29 | 山东大学 | Method for generating network based on optical flow, image and motion confrontation to improve recognition accuracy rate of anxiety, depression and angry expression |
| KR20190084913A (en) * | 2019-06-28 | 2019-07-17 | 엘지전자 주식회사 | Refrigerator and method for operating the refrigerator |
| CN110334768B (en) * | 2019-07-08 | 2023-08-22 | 创新奇智(重庆)科技有限公司 | Refrigerator display detection method and system and electronic equipment |
| CN110163828B (en) * | 2019-07-10 | 2022-02-18 | 复旦大学 | Mammary gland calcification image optimization system and method based on ultrasonic radio frequency signals |
| CN110363159A (en) * | 2019-07-17 | 2019-10-22 | 腾讯科技（深圳）有限公司 | Image-recognizing method, device, electronic equipment and computer readable storage medium |
| CN110378288B (en) * | 2019-07-19 | 2021-03-26 | 合肥工业大学 | Deep learning-based multi-stage space-time moving target detection method |
| US20220326384A1 (en) * | 2019-08-23 | 2022-10-13 | Semiconductor Energy Laboratory Co., Ltd. | Imaging device, distance estimation device, and moving object |
| CN110715401B (en) * | 2019-08-29 | 2020-12-15 | 珠海格力电器股份有限公司 | Defrosting control method and device for air conditioning equipment, medium and air conditioning equipment |
| CN110837776A (en) * | 2019-10-09 | 2020-02-25 | 广东工业大学 | Pulse neural network handwritten Chinese character recognition method based on STDP |
| CN111008726B (en) * | 2019-10-28 | 2023-08-29 | 武汉理工大学 | Class picture conversion method in power load prediction |
| CN110850717B (en) * | 2019-11-19 | 2022-05-20 | 浙江工业大学 | Neural network heat pump defrosting control device and control method by utilizing fan current |
| CN110853032B (en) * | 2019-11-21 | 2022-11-01 | 北京航空航天大学 | Unmanned aerial vehicle video tag acquisition method based on multi-mode deep learning |
| CN111156657A (en) * | 2019-12-25 | 2020-05-15 | 珠海格力电器股份有限公司 | Air conditioner frosting state determining method and device |
| CN111059847A (en) * | 2019-12-30 | 2020-04-24 | 南京工程学院 | Evaporator frosting visual detection system and detection method thereof |
| CN111199543A (en) * | 2020-01-07 | 2020-05-26 | 南京航空航天大学 | Refrigerator-freezer surface defect detects based on convolutional neural network |
| CN111353390A (en) * | 2020-01-17 | 2020-06-30 | 道和安邦（天津）安防科技有限公司 | Micro-expression recognition method based on deep learning |
| CN111383214B (en) * | 2020-03-10 | 2021-02-19 | 长沙慧维智能医疗科技有限公司 | Real-time endoscope enteroscope polyp detection system |
| CN111428676B (en) * | 2020-04-01 | 2023-04-07 | 南京信息工程大学 | Short-term rainfall prediction method based on sparse correspondence and deep neural network |
| CN111507929A (en) * | 2020-04-15 | 2020-08-07 | 上海眼控科技股份有限公司 | Meteorological cloud picture prediction method and device, computer equipment and storage medium |
| CN111651941B (en) * | 2020-04-30 | 2022-05-17 | 北京航空航天大学 | Global ionosphere electron total content prediction algorithm |
| CN111784633B (en) * | 2020-05-26 | 2024-02-06 | 西安理工大学 | Insulator defect automatic detection algorithm for electric power inspection video |
| CN111724408B (en) * | 2020-06-05 | 2021-09-03 | 广东海洋大学 | Verification experiment method of abnormal driving behavior algorithm model based on 5G communication |
| CN111639719B (en) * | 2020-06-08 | 2023-04-07 | 安徽大学 | Footprint image retrieval method based on space-time motion and feature fusion |
| CN111814611B (en) * | 2020-06-24 | 2022-09-13 | 重庆邮电大学 | Multi-scale face age estimation method and system embedded with high-order information |
| CN112507785A (en) * | 2020-11-02 | 2021-03-16 | 北京工商大学 | Rolling bearing fault analysis based on CNN and LSTM |
| CN112580777A (en) * | 2020-11-11 | 2021-03-30 | 暨南大学 | Attention mechanism-based deep neural network plug-in and image identification method |
| CN112529934B (en) * | 2020-12-02 | 2023-12-19 | 北京航空航天大学杭州创新研究院 | Multi-target tracking method, device, electronic equipment and storage medium |
| CN112464844A (en) * | 2020-12-07 | 2021-03-09 | 天津科技大学 | Human behavior and action recognition method based on deep learning and moving target detection |
| CN112906787B (en) * | 2021-02-10 | 2023-09-15 | 宁波诺丁汉新材料研究院有限公司 | Industrial boiler fault identification method and system |
| CN112906631B (en) * | 2021-03-17 | 2022-07-29 | 南京邮电大学 | Dangerous driving behavior detection method and detection system based on video |
-
2021
- 2021-06-29 CN CN202110727931.1A patent/CN113536989B/en active Active
Non-Patent Citations (2)
| Title |
|---|
| 基于神经网络的空气源热泵除霜诊断模型;曹亮等;《暖通空调》;20060630(第06期);100-103 * |
| 多核并行脉冲神经网络模拟器的设计;陈靖宇等;《计算机工程与应用》;20201115;第56卷(第22期);244-250 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113536989A (en) | 2021-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109712127B (en) | Power transmission line fault detection method for machine inspection video stream | |
| CN113469050B (en) | Flame detection method based on image fine classification | |
| CN113449660B (en) | Abnormal event detection method of space-time variation self-coding network based on self-attention enhancement | |
| CN114484731A (en) | Method and device for diagnosing faults of central air conditioner based on stacking fusion algorithm | |
| CN114120093B (en) | Coal gangue target detection method based on improved YOLOv algorithm | |
| CN113435593B (en) | Refrigeration equipment frosting prediction method based on sensor time sequence data analysis | |
| CN113569756B (en) | Abnormal behavior detection and positioning method, system, terminal equipment and readable storage medium | |
| Chen et al. | Unsupervised curriculum domain adaptation for no-reference video quality assessment | |
| CN113536989B (en) | Refrigerator frosting monitoring method and system based on frame-by-frame analysis of camera video | |
| CN111191610A (en) | People flow detection and processing method in video monitoring | |
| CN111339892B (en) | Swimming pool drowning detection method based on end-to-end 3D convolutional neural network | |
| CN114821155A (en) | Multi-label classification method and system based on deformable NTS-NET neural network | |
| Guo et al. | Grape leaf disease detection based on attention mechanisms | |
| CN117541534A (en) | Power transmission line inspection method based on unmanned plane and CNN-BiLSTM model | |
| CN117579513B (en) | Visual operation and maintenance system and method for convergence and diversion equipment | |
| CN112861443B (en) | Advanced learning fault diagnosis method integrated with priori knowledge | |
| CN107527363A (en) | A kind of cold storage plant deposits object storage management system and cold storage plant | |
| CN107527060A (en) | A kind of cold storage plant deposits object storage management system and cold storage plant | |
| CN114627286A (en) | Method for detecting wagon staff invasion based on PSPNet and improved YOLOv4 | |
| CN116977807A (en) | Multi-sensor fusion-based intelligent monitoring system and method for refrigerator | |
| CN116994161A (en) | Insulator defect detection method based on improved YOLOv5 | |
| Pillai et al. | Fine-Tuned EfficientNetB4 Transfer Learning Model for Weather Classification | |
| CN115527151A (en) | Video anomaly detection method and system, electronic equipment and storage medium | |
| CN113469228A (en) | Power load abnormal value identification method based on data flow space-time characteristics | |
| CN115412451A (en) | Internet link utilization rate prediction method based on fusion model |
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 |