CN109145830A - A kind of intelligence water gauge recognition methods - Google Patents
A kind of intelligence water gauge recognition methods Download PDFInfo
- Publication number
- CN109145830A CN109145830A CN201810974283.8A CN201810974283A CN109145830A CN 109145830 A CN109145830 A CN 109145830A CN 201810974283 A CN201810974283 A CN 201810974283A CN 109145830 A CN109145830 A CN 109145830A
- Authority
- CN
- China
- Prior art keywords
- water gauge
- water
- layer
- neural network
- image
- 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.)
- Granted
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 168
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000012549 training Methods 0.000 claims abstract description 31
- 238000003062 neural network model Methods 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims abstract description 6
- 238000000605 extraction Methods 0.000 claims abstract 2
- 238000013528 artificial neural network Methods 0.000 claims description 18
- 238000012360 testing method Methods 0.000 claims description 8
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 210000002364 input neuron Anatomy 0.000 claims description 6
- 210000004205 output neuron Anatomy 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 2
- 238000011478 gradient descent method Methods 0.000 claims description 2
- 210000002569 neuron Anatomy 0.000 claims description 2
- 238000011176 pooling Methods 0.000 claims 8
- 230000003190 augmentative effect Effects 0.000 claims 1
- 230000007423 decrease Effects 0.000 claims 1
- 239000011159 matrix material Substances 0.000 claims 1
- 238000010200 validation analysis Methods 0.000 claims 1
- 238000012544 monitoring process Methods 0.000 abstract description 15
- 238000012795 verification Methods 0.000 abstract 1
- 230000003416 augmentation Effects 0.000 description 3
- 239000012141 concentrate Substances 0.000 description 3
- 230000001537 neural effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013434 data augmentation Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01F—MEASURING VOLUME, VOLUME FLOW, MASS FLOW OR LIQUID LEVEL; METERING BY VOLUME
- G01F23/00—Indicating or measuring liquid level or level of fluent solid material, e.g. indicating in terms of volume or indicating by means of an alarm
- G01F23/04—Indicating or measuring liquid level or level of fluent solid material, e.g. indicating in terms of volume or indicating by means of an alarm by dip members, e.g. dip-sticks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
- G06V10/449—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Biodiversity & Conservation Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Fluid Mechanics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
本发明公开了一种智能水尺识别方法，属于水位监测技术领域，包括：1)采集水尺图像，并将所有水尺图像分为训练集和验证集；筛选出训练集中的水尺图像进行提取，得到水尺的目标区域；2)设水尺的宽度为a，采用长为2a，宽为2a的矩形以一定步长截取目标区域，截取的过程中保证截取下的部分不超过水尺的左右边缘；3)将截取的水面以上的图像作为正样本，水面以下的图像作为负样本，存入对应的水尺刻度值的数据集中；4)利用所有水尺刻度值对应的数据集训练神经网络模型；5)截取待识别水尺图像的目标区域，经过步骤2)处理后使用训练好的神经网络模型判断水位，并将图像存入对应的水尺刻度值的数据集中。能对水位线不清晰的图片进行高效识别。
The invention discloses an intelligent water gauge identification method, belonging to the technical field of water level monitoring, comprising: 1) collecting water gauge images, and dividing all water gauge images into a training set and a verification set; Extraction to obtain the target area of the water gauge; 2) Set the width of the water gauge as a, use a rectangle with a length of 2a and a width of 2a to intercept the target area with a certain step length, and ensure that the intercepted part does not exceed the water gauge during the interception process. 3) Take the intercepted images above the water surface as positive samples, and the images below the water surface as negative samples, and store them in the data set of the corresponding water gauge scale values; 4) Use all data sets corresponding to the water gauge scale values for training Neural network model; 5) Intercept the target area of the water gauge image to be identified, use the trained neural network model to judge the water level after processing in step 2), and store the image in the data set of the corresponding water gauge scale value. It can efficiently identify pictures with unclear water level lines.
Description
Technical field
The present invention relates to water level monitoring technical fields, specifically, being related to a kind of intelligent water gauge recognition methods.
Background technique
Water level monitoring is of great significance for the important monitoring index of the water bodys such as river, river, reservoir.In the prior art,
Conventional water level monitoring method has sensor monitoring and water-level measuring post personal monitoring.Wherein, water-saving to reach in terms of water gauge identification
Purpose, using the method for video image monitoring to river, irrigate the water level in canal and monitor in real time.It is read although with artificial
The method of video can recorde the data such as the water level of water gauge, but due to monitoring point enormous amount, recording personnel's attention cannot grow
Time concentrates, and in face of so many monitoring image, often due to the carelessness of record personnel, input when may be generated
Situations such as fault.There are also the methods of many automatic identification water gauges, such as:
The Chinese patent literature that publication No. is CN101886942A discloses a kind of water gauge identification method and water level detecting side
Method, this method shoot water gauge using video image, then achieve the purpose that detection, but the party by the identification to image
Method is needed using specific scale, and application range is very limited.
In addition, publication No. be CN102975826A Chinese patent literature disclose it is a kind of based on the portable of machine vision
Shipping depth gauge detects automatically and recognition methods, and this method uses machine vision and image processing techniques, to ship under offshore environment
Water gauge video data is handled, and detects and identify shipping depth gauge scale, and count to the testing result of video successive frame
Analysis, finally obtains drauht value.Publication No. is that the Chinese patent literature of CN105046212A discloses a kind of waterline scale
Automatic identifying method, this method obtain the packet in water surface line of demarcation or more by the water surface line of demarcation in criterion of identification sample image
The water gauge digital picture of aqueous footage word obtains water gauge scale into identification.And publication No. is the China of CN108318101A
Patent document discloses a kind of water gauge water level video intelligent monitoring method based on deep learning algorithm and system, this method include
Video acquisition, video frame processing, water level line identification and water level measuring and calculating and etc..It is supervised using deep learning neural fusion water level
The intelligence and automation of survey.
Although above method is able to achieve the automation of water level monitoring without specific scale, water level is being carried out
It only considered the opaque situation of water turbidity during monitoring, when water quality is limpid, the color and water level line of water are not easy to know
Bigger error is had when other, therefore use scope is caused to be restricted.
Summary of the invention
It is an object of the present invention to provide a kind of intelligent water gauge recognition methods, this method during carrying out water level monitoring,
The opaque situation of water turbidity can be not only identified, water quality is limpid, color of water and water level line do not allow situation easy to identify
Under also can be carried out efficient identification.
To achieve the goals above, intelligent water gauge recognition methods provided by the invention the following steps are included:
1) water gauge image is acquired, and all water gauge images are divided into training set and verifying collection；Filter out the water in training set
Ruler image extracts, and obtains the target area of water gauge；
2) width of water gauge is set as a, using a length of 2a, width intercepts the target area for the rectangle of 2a with a fixed step size,
The part under guaranteeing to intercept is needed to be no more than the left and right edges of water gauge during interception；
3) using the image more than water surface of interception as positive sample, water surface image below is corresponded to as negative sample, deposit
Water gauge scale value data set in；
4) the corresponding data set training neural network model of all water gauge scale values is utilized；
5) target area for intercepting water gauge image to be identified uses trained neural network mould after step 2) processing
Type judges water level, and image is stored in the data set of corresponding water gauge scale value.
In above-mentioned technical proposal, using video image automatic identification, self registering method, artificial work can be not only reduced
It measures, and accurately and timely can be automatically stored and be shown.Since the amount of images of water gauge is relatively fewer, using random
The method for sampling small rectangle can be repeated as many times and utilize same image, and guarantee the diversity of sample, so that the model of training
Robustness is stronger.
Specifically, step 1) further include: random augmentation is carried out to the water gauge image of acquisition, and will be schemed using bilinear interpolation
Piece is adjusted to fixed size, and image is normalized.
Carrying out random augmentation to picture includes at random plus dry, random brightness adjustment, random cropping, random mirror image reversal etc.；
Image normalization can speed the speed of neural metwork training.
Specifically, the color characteristic in step 1) by water gauge determines the target area of water gauge；If there are multiple regions, lead to
The size in region is crossed, length-width ratio is screened；If still there is multiple regions, calculate in each region the mean value and variance of pixel with
Water gauge region in training set compares, and selects immediate region.
When actually identifying water gauge water level, water gauge image needs to be tested are placed on verifying and concentrate.Confirm water gauge region
Method be mainly reading by the color characteristic of water gauge, on water gauge and scale is blue or red, this color will be met
The region of feature scans for.
Specifically, step 3) further include: the positive sample and the negative sample are divided into instruction respectively with the ratio of 1:1
Practice collection and test set, and the corresponding water gauge scale value of truncated picture is converted into binary file storage, convenient for training
When data reading.
Specifically, include: using the method for data set training neural network model in step 4)
4-1) weight of convolutional layer and full articulamentum is initialized using xavier initialization, while setting super ginseng
Number；The purpose of initialization is to allow neural network to acquire useful information in study, and hyper parameter includes learning rate, batch_
Size, epoch etc.；
4-2) picture in data set is put into convolutional neural networks and is trained, which successively includes
Output layer, the first convolutional layer, the first pond layer, the second convolutional layer, the second pond layer, the first full articulamentum, the second full articulamentum
With Loss layers, output result be 2 neurons；
Loss 4-3) is calculated using softmax function, then gradient is calculated by gradient descent method, uses back-propagation algorithm
Disease gradient updates the parameter of neural network；
4-4) train neural network until the loss of test set no longer declines.
Specifically, step 4-2) in input layer input be size be 20x20x3 RGB three-dimensional matrice；First convolutional layer
Convolution kernel size be 5x5, step-length 1, dimension 3x96；The size of the convolution kernel of second convolutional layer is 3x3, step-length 1, dimension
Degree is 96x256；The size of the Chi Huahe of first pond layer is 2x2, and step-length 2, the feature sizes of Chi Huahou are 10x10x96；
The size of the Chi Huahe of second pond layer is 2x2, and step-length 2, the feature sizes of Chi Huahou are 5x5x256；First full articulamentum
Input neuron be 6400, output neuron be 512, drop out be 0.5；The input neuron of second articulamentum is
512, output neuron is 2.Specifically, step 4-4) in training neural network when, first use high
Learning rate when training tends towards stability to loss, is declined 10 times, continues to train, and repeat aforesaid operations twice by habit rate.
Specifically, the method for water level is judged in step 5) using neural network model are as follows:
It sets the width of the target area of water gauge image to be identified 5-1) as b, target area is divided into m a length of b, width is
The fritter of b is sent among neural network after each fritter is adjusted to fixed size, judges the region by the output of full articulamentum
Whether under water；
It 5-2) finds from waterborne and is converted into two underwater critical zones, then the two regions are divided into n a length of c, it is wide
For the fritter of b, wherein n*c=2b；
It is put into after fritter 5-3) is adjusted to the input size of neural network in order from top to bottom, passes through full articulamentum
Whether under water output judge the region, when it is underwater for exporting result, the Y coordinate of fritter at this time is recorded, according to coordinate
Calculate the height of water level.
Step 5-1) it is used as coarse adjustment, can quickly position the Position Approximate of water level line, claim 5-2)~5-3) it is thin
It adjusts, keeps the position of water level line more accurate.
Specifically, c takes b/8.Improve the accuracy of identification.
Compared with prior art, the invention has the benefit that
The accuracy of intelligent water gauge recognition methods in example of the present invention is higher, strong antijamming capability, to environment and camera picture
The requirement of element is lower.It can not be illuminated by the light intensity substantially using the water level identification of water gauge in outdoor environment, leaf blocks,
The influence of the natural causes such as rainy.Even if camera distance equally can be identified precisely farther out, color limpid for water quality, water
Situation easy to identify is not allowed with water level line, can be carried out efficient identification yet.
Detailed description of the invention
Fig. 1 is the flow chart of the intelligent water gauge recognition methods of the embodiment of the present invention；
Fig. 2 is water gauge picture to be identified in the embodiment of the present invention；
Fig. 3 is the water gauge picture extracted in the embodiment of the present invention；
Fig. 4 is the picture of water gauge to be identified in the embodiment of the present invention；
Fig. 5 is the picture extracted in the embodiment of the present invention as data set positive sample；
Fig. 6 is the picture extracted in the embodiment of the present invention as data set negative sample；
Fig. 7 is convolutional neural networks of embodiment of the present invention structure chart；
Fig. 8 is water gauge picture to be identified when the water surface is more muddy of the embodiment of the present invention；
Fig. 9 is water gauge picture to be identified when the water surface is more limpid of the embodiment of the present invention；
Figure 10 is the picture of water gauge to be identified when the water surface of the embodiment of the present invention is very limpid.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, with reference to embodiments and its attached drawing is to this hair
It is bright to be described further.
Embodiment
Referring to fig. 2, after the specific image of certain water level to be identified is extracted as shown in Figure 3.It is needed when processing by picture point
, with underwater two parts, to calculate water level at that time on the water surface further according to the position of water level line, reaching real-time automatic monitoring water
The purpose of position.
Referring to Fig. 1, the intelligent water gauge recognition methods of the present embodiment the following steps are included:
Picture in the database of water gauge is divided into training set to S1 and verifying collects, and filters out from training set relatively clear
Picture, as shown in figure 4, the width of water gauge be a.With a length of 2a, width is that the rectangle of 2a size is intercepted at random from figure in the water surface
The water gauge of top guarantees that the part under interception is no more than the left and right edges of water gauge, and not comprising underwater portion, as shown in Figure 5.It will
Positive sample of the picture intercepted as data set.
S2 selects picture at random from training set, and equally with a length of 2a, width is that the rectangle of 2a intercepts water gauge underwater at random
Picture, as shown in Figure 6.Using picture as the negative sample of data set.
Positive negative sample is divided into training set and test set respectively with the ratio of 1:1 by S3, and image and label are converted to two
Binary file storage, convenient for the reading of data when training.
S4 reads image data from data set, carries out random augmentation to picture, including at random plus dry, random brightness
Adjustment, random cropping, random mirror image reversal etc..Picture is adjusted to fixed size using bilinear interpolation.Image is returned
One changes, and image normalization can speed the speed of neural metwork training.
S5 using xavier initialization the weight of convolutional layer and full articulamentum is initialized, the purpose of initialization be for
The neural network is allowed to acquire useful information in study.Hyper parameter is set simultaneously, including learning rate, batch_
Size, epoch etc..
The picture handled well is put into convolutional neural networks and is trained by S6.The construction of the convolutional neural networks such as Fig. 7
Shown, for the data set of water level identification, the convolutional neural networks model that this example uses has eight layers (comprising input and output layer) altogether.
Wherein first layer is input layer, and input is RGB three-dimensional matrice that size is 20x20x3.Before training, first by input picture into
Row random cropping, at random plus dry, random mirror face turning, input of the renormalization to [0,1] as network.The second layer is convolution
The convolution kernel size of layer, the convolutional layer is 5x5, and the weight of step-length 1, dimension 3x96, convolution kernel is initialized using xavier,
Bias term is 0.The feature sizes obtained after convolution are 20x20x96.Third layer is pond layer, and using maximum pond, Chi Huahe's is big
Small is 2x2, step-length 2.Feature sizes behind pond are 10x10x96.4th layer is convolutional layer, and the size of convolution kernel is 3x3,
Step-length is 1, and the weight of dimension 96x256, convolution kernel are initialized using xavier, bias term 0.Feature sizes after convolution
For 10x10x256.Layer 5 is pond layer, and using maximum pond, the size of Chi Huahe is 2x2, step-length 2.Feature behind pond
Size is 5x5x256.Layer 6 is full articulamentum, and input neuron is 6400, and output neuron is 512, drop out
It is 0.5.Layer 7 is full articulamentum, and input neuron is 512, and output neuron is 2.8th layer is Loss layers, is used
SOFTMAX calculates loss, and stochastic gradient descent carries out backpropagation.
S7 training neural network no longer declines until the loss of test set.
S8 is concentrated from verifying and is chosen picture as shown in Figure 8, and the region where water gauge is determined by the color characteristic of water gauge.
S9 sets the width in region as b, and by region segmentation at m a length of b, width is the fritter of b.Fritter is adjusted to fixed big
It is sent among neural network after small, whether judge that the region belongs to by the output of full articulamentum is under water.
S10 finds from waterborne and is converted into two underwater critical zones, then the two regions are divided into n a length of c, and width is
The fritter of b, wherein n*c=2b.It is put into after fritter is adjusted to the input size of neural network in order from top to bottom, by complete
Whether it is under water that the output of articulamentum judges that the region belongs to.When it is underwater for exporting result, by the Y coordinate of fritter at this time
Record, the height of water level is calculated according to coordinate.
This example uses data augmentation method, joined more random factors, so that model anti-interference ability is reinforced, and not
It is easy over-fitting.Neural network structure in this example can according to need adjustment, increase the number of plies of network and the dimension of convolution kernel
The effect of classification can be enhanced.In this example, when training neural network, a higher learning rate, training are used first
When tending towards stability to loss, learning rate is declined 10 times, continues to train, repeats aforesaid operations twice, it is available more satisfactory
Result.
If identifying the water level value of picture shown in Fig. 9, Fig. 9 is chosen, repeats step S8 to step S10, Figure 10 also one
Sample.When recognition result shows the water gauge figure such as Fig. 8, the water surface is more muddy, and discrimination is higher up and down for the water surface, and error can be in 1cm
Within；As Fig. 9 water gauge figure when, the water surface is more limpid, and error is generally in 1~3cm；As Figure 10 water gauge figure when, the water surface is very
When limpid, it is easy to produce biggish error, generally in 3~5cm.It should be noted that this result is that due to currently scheming
Caused by sheet data amount is also considerably less, when image data reaches certain quantity, then its precision can greatly improve.Compared to
The method of others image recognition water gauge at present, precision of the invention is high, and strong antijamming capability wants environment and camera pixel
Ask low.
When actually identifying water gauge water level, the image needs of water gauge to be tested are placed on verifying and concentrate.Water is confirmed in step S8
The method of ruler region is mainly the color characteristic by water gauge, and the reading and scale on water gauge are blue or red, will
The region for meeting this color characteristic in step S8 scans for.If there is multiple regions, by the size in region, length-width ratio is carried out
Screening.If still there are multiple regions, calculates the mean value with variance of pixel in each region and oppose with the water gauge region in training set
Than selecting immediate region.This example obtains the coordinate of water level line using two steps of step S9 and step S10, wherein walking
Rapid S9 can quickly position the Position Approximate of water level line as coarse adjustment, and step S10 is fine tuning, and speed is slower, but is positioned more
Precisely.Wherein the accuracy of step S10 is influenced by the size of c, and c takes effect when b/8 preferable in test.
Claims (9)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810974283.8A CN109145830B (en) | 2018-08-24 | 2018-08-24 | Intelligent water gauge identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810974283.8A CN109145830B (en) | 2018-08-24 | 2018-08-24 | Intelligent water gauge identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109145830A true CN109145830A (en) | 2019-01-04 |
| CN109145830B CN109145830B (en) | 2020-08-25 |
Family
ID=64827875
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810974283.8A Expired - Fee Related CN109145830B (en) | 2018-08-24 | 2018-08-24 | Intelligent water gauge identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109145830B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109815865A (en) * | 2019-01-11 | 2019-05-28 | 江河瑞通（北京）技术有限公司 | A kind of water level recognition methods and system based on virtual water gauge |
| CN109870211A (en) * | 2019-01-22 | 2019-06-11 | 珠海格力电器股份有限公司 | Method and device for detecting rice amount in rice storage equipment and rice barrel |
| CN110033037A (en) * | 2019-04-08 | 2019-07-19 | 重庆邮电大学 | A kind of recognition methods of digital instrument reading |
| CN110223341A (en) * | 2019-06-14 | 2019-09-10 | 北京国信华源科技有限公司 | A kind of Intelligent water level monitoring method based on image recognition |
| CN110472636A (en) * | 2019-07-26 | 2019-11-19 | 四创科技有限公司 | Water gauge E font scale recognition methods based on deep learning |
| CN111292371A (en) * | 2020-02-21 | 2020-06-16 | 廖赟 | Intelligent water level detection equipment, intelligent water level monitoring management device and detection method |
| CN111488846A (en) * | 2020-04-16 | 2020-08-04 | 上海芯翌智能科技有限公司 | Method and equipment for identifying water level |
| CN112013921A (en) * | 2019-05-30 | 2020-12-01 | 杭州海康威视数字技术股份有限公司 | Method, device and system for acquiring water level information based on water level gauge measurement image |
| CN112183470A (en) * | 2020-10-28 | 2021-01-05 | 长江大学 | Ship water gauge identification method and equipment and storage medium |
| CN112465061A (en) * | 2020-12-10 | 2021-03-09 | 厦门四信通信科技有限公司 | Water level identification method, device, equipment and storage medium |
| CN112561876A (en) * | 2020-12-14 | 2021-03-26 | 中南大学 | Image-based pond and reservoir water quality detection method and system |
| CN113159005A (en) * | 2021-05-28 | 2021-07-23 | 青海中水数易信息科技有限责任公司 | Machine learning-based water level and foreign matter identification integrated monitoring system and method |
| CN113221898A (en) * | 2021-04-16 | 2021-08-06 | 北京科技大学 | Automatic water level gauge reading method |
| CN113252614A (en) * | 2021-05-26 | 2021-08-13 | 浙江大学 | Transparency detection method based on machine vision |
| CN113822104A (en) * | 2020-07-07 | 2021-12-21 | 湖北亿立能科技股份有限公司 | Artificial intelligence water surface detection system based on virtual scale of many candidates |
| CN114926786A (en) * | 2022-06-22 | 2022-08-19 | 国能黄骅港务有限责任公司 | Ship water gauge tracking method and device, storage medium and electronic equipment |
| CN115546235A (en) * | 2022-11-23 | 2022-12-30 | 成都见海科技有限公司 | Water level identification method and system based on image segmentation and storage medium |
| CN117974549A (en) * | 2023-12-11 | 2024-05-03 | 广西壮族自治区水利科学研究院 | A water level calculation method based on image recognition and dual-region convolutional neural network |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050197981A1 (en) * | 2004-01-20 | 2005-09-08 | Bingham Clifton W. | Method for identifying unanticipated changes in multi-dimensional data sets |
| CN102975826A (en) * | 2012-12-03 | 2013-03-20 | 上海海事大学 | Portable ship water gauge automatic detection and identification method based on machine vision |
| CN103902972A (en) * | 2014-03-21 | 2014-07-02 | 哈尔滨工程大学 | Water surface moving platform visual system image analyzing and processing method |
| CN106408566A (en) * | 2016-11-10 | 2017-02-15 | 深圳大学 | Fetal ultrasound image quality control method and system |
| CN106557764A (en) * | 2016-11-02 | 2017-04-05 | 江西理工大学 | A kind of water level recognition methodss based on binary-coded character water gauge and image procossing |
| CN106803090A (en) * | 2016-12-05 | 2017-06-06 | 中国银联股份有限公司 | A kind of image-recognizing method and device |
| CN108073905A (en) * | 2017-12-21 | 2018-05-25 | 北京邮电大学 | A kind of method, system and the equipment of intelligence water gauge reading |
| CN108318101A (en) * | 2017-12-26 | 2018-07-24 | 北京市水利自动化研究所 | Water gauge water level video intelligent monitoring method based on deep learning algorithm and system |
| CN108423136A (en) * | 2018-02-27 | 2018-08-21 | 力鸿智信（北京）科技有限公司 | A kind of ship pay load determines method and device |
-
2018
- 2018-08-24 CN CN201810974283.8A patent/CN109145830B/en not_active Expired - Fee Related
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050197981A1 (en) * | 2004-01-20 | 2005-09-08 | Bingham Clifton W. | Method for identifying unanticipated changes in multi-dimensional data sets |
| CN102975826A (en) * | 2012-12-03 | 2013-03-20 | 上海海事大学 | Portable ship water gauge automatic detection and identification method based on machine vision |
| CN103902972A (en) * | 2014-03-21 | 2014-07-02 | 哈尔滨工程大学 | Water surface moving platform visual system image analyzing and processing method |
| CN106557764A (en) * | 2016-11-02 | 2017-04-05 | 江西理工大学 | A kind of water level recognition methodss based on binary-coded character water gauge and image procossing |
| CN106408566A (en) * | 2016-11-10 | 2017-02-15 | 深圳大学 | Fetal ultrasound image quality control method and system |
| CN106803090A (en) * | 2016-12-05 | 2017-06-06 | 中国银联股份有限公司 | A kind of image-recognizing method and device |
| CN108073905A (en) * | 2017-12-21 | 2018-05-25 | 北京邮电大学 | A kind of method, system and the equipment of intelligence water gauge reading |
| CN108318101A (en) * | 2017-12-26 | 2018-07-24 | 北京市水利自动化研究所 | Water gauge water level video intelligent monitoring method based on deep learning algorithm and system |
| CN108423136A (en) * | 2018-02-27 | 2018-08-21 | 力鸿智信（北京）科技有限公司 | A kind of ship pay load determines method and device |
Non-Patent Citations (3)
| Title |
|---|
| TEGUH TRIANTORO 等: "Image Based Water Gauge Reading Developed with ANN Kohonen", 《2014 IEEE INTERNATIONAL CONFERENCE ON ELECTRICAL ENGINEERING AND COMPUTER SCIENCE》 * |
| 尤良权: "基于二进制编码水尺和数字图像的水位远程监控系统设计", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王丽君 等: "基于卷积神经网络的位置识别", 《电子科技》 * |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109815865A (en) * | 2019-01-11 | 2019-05-28 | 江河瑞通（北京）技术有限公司 | A kind of water level recognition methods and system based on virtual water gauge |
| CN109815865B (en) * | 2019-01-11 | 2021-11-16 | 江河瑞通（北京）技术有限公司 | Water level identification method and system based on virtual water gauge |
| CN109870211A (en) * | 2019-01-22 | 2019-06-11 | 珠海格力电器股份有限公司 | Method and device for detecting rice amount in rice storage equipment and rice barrel |
| CN109870211B (en) * | 2019-01-22 | 2020-07-28 | 珠海格力电器股份有限公司 | Method and device for detecting rice amount in rice storage equipment and rice barrel |
| CN110033037A (en) * | 2019-04-08 | 2019-07-19 | 重庆邮电大学 | A kind of recognition methods of digital instrument reading |
| CN112013921B (en) * | 2019-05-30 | 2023-06-23 | 杭州海康威视数字技术股份有限公司 | Method, device and system for acquiring water level information based on water level gauge measurement image |
| CN112013921A (en) * | 2019-05-30 | 2020-12-01 | 杭州海康威视数字技术股份有限公司 | Method, device and system for acquiring water level information based on water level gauge measurement image |
| CN110223341A (en) * | 2019-06-14 | 2019-09-10 | 北京国信华源科技有限公司 | A kind of Intelligent water level monitoring method based on image recognition |
| CN110223341B (en) * | 2019-06-14 | 2024-05-28 | 北京国信华源科技有限公司 | Intelligent water level monitoring method based on image recognition |
| CN110472636A (en) * | 2019-07-26 | 2019-11-19 | 四创科技有限公司 | Water gauge E font scale recognition methods based on deep learning |
| CN110472636B (en) * | 2019-07-26 | 2022-10-14 | 四创科技有限公司 | Deep learning-based water gauge E-shaped scale identification method |
| CN111292371A (en) * | 2020-02-21 | 2020-06-16 | 廖赟 | Intelligent water level detection equipment, intelligent water level monitoring management device and detection method |
| CN111488846A (en) * | 2020-04-16 | 2020-08-04 | 上海芯翌智能科技有限公司 | Method and equipment for identifying water level |
| CN113822104A (en) * | 2020-07-07 | 2021-12-21 | 湖北亿立能科技股份有限公司 | Artificial intelligence water surface detection system based on virtual scale of many candidates |
| CN113822104B (en) * | 2020-07-07 | 2023-11-03 | 湖北亿立能科技股份有限公司 | Artificial intelligence surface of water detecting system based on virtual scale of many candidates |
| CN112183470B (en) * | 2020-10-28 | 2022-07-19 | 长江大学 | Ship water gauge identification method and equipment and storage medium |
| CN112183470A (en) * | 2020-10-28 | 2021-01-05 | 长江大学 | Ship water gauge identification method and equipment and storage medium |
| CN112465061A (en) * | 2020-12-10 | 2021-03-09 | 厦门四信通信科技有限公司 | Water level identification method, device, equipment and storage medium |
| CN112465061B (en) * | 2020-12-10 | 2023-03-24 | 厦门四信通信科技有限公司 | Water level identification method, device, equipment and storage medium |
| CN112561876A (en) * | 2020-12-14 | 2021-03-26 | 中南大学 | Image-based pond and reservoir water quality detection method and system |
| CN112561876B (en) * | 2020-12-14 | 2024-02-23 | 中南大学 | Image-based water quality detection method and system for ponds and reservoirs |
| CN113221898A (en) * | 2021-04-16 | 2021-08-06 | 北京科技大学 | Automatic water level gauge reading method |
| CN113252614A (en) * | 2021-05-26 | 2021-08-13 | 浙江大学 | Transparency detection method based on machine vision |
| CN113159005A (en) * | 2021-05-28 | 2021-07-23 | 青海中水数易信息科技有限责任公司 | Machine learning-based water level and foreign matter identification integrated monitoring system and method |
| CN114926786A (en) * | 2022-06-22 | 2022-08-19 | 国能黄骅港务有限责任公司 | Ship water gauge tracking method and device, storage medium and electronic equipment |
| CN115546235A (en) * | 2022-11-23 | 2022-12-30 | 成都见海科技有限公司 | Water level identification method and system based on image segmentation and storage medium |
| CN117974549A (en) * | 2023-12-11 | 2024-05-03 | 广西壮族自治区水利科学研究院 | A water level calculation method based on image recognition and dual-region convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109145830B (en) | 2020-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109145830A (en) | A kind of intelligence water gauge recognition methods | |
| CN113392775B (en) | Sugarcane seedling automatic identification and counting method based on deep neural network | |
| CN109460753B (en) | Method for detecting floating object on water | |
| CN110414334B (en) | An intelligent water quality identification method based on drone inspection | |
| CN109522889A (en) | A Method for Recognition and Estimation of Hydrometer Water Level Based on Image Analysis | |
| CN109285139A (en) | A deep learning-based X-ray imaging weld inspection method | |
| CN111598098B (en) | A fully convolutional neural network-based method for water level detection and validity identification of water gauge | |
| CN113435460A (en) | Method for identifying brilliant particle limestone image | |
| KR101743270B1 (en) | The method for separating and recognizing individual plankton using deep learning in a clumped or dispersed microscopic plankton image | |
| CN108090906A (en) | A kind of uterine neck image processing method and device based on region nomination | |
| CN111462044B (en) | Greenhouse strawberry detection and maturity evaluation method based on deep learning model | |
| CN108564085A (en) | A kind of method of automatic reading pointer type instrument reading | |
| CN115861170A (en) | Surface defect detection method based on improved YOLO V4 algorithm | |
| CN112418028A (en) | Satellite image ship identification and segmentation method based on deep learning | |
| CN115512215B (en) | Underwater biological monitoring method, device and storage medium | |
| CN106874929A (en) | A kind of pearl sorting technique based on deep learning | |
| CN111339902A (en) | Liquid crystal display number identification method and device of digital display instrument | |
| CN107016353A (en) | A kind of method and system of variable resolution target detection and identification integration | |
| CN111141653A (en) | Tunnel leakage rate prediction method based on neural network | |
| CN113469097A (en) | SSD (solid State disk) network-based real-time detection method for water surface floating object multiple cameras | |
| Wang et al. | Vision-based in situ monitoring of plankton size spectra via a convolutional neural network | |
| CN112184627A (en) | Citrus fresh-keeping quality detection method based on image processing and neural network and application | |
| CN115019181B (en) | Remote sensing image rotating target detection method, electronic equipment and storage medium | |
| CN118351100A (en) | Image definition detection and processing method based on deep learning and gradient analysis | |
| CN117830826A (en) | Concrete dam surface slender crack detection method based on unmanned aerial vehicle |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200825 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |