CN109376740A - A kind of water gauge reading detection method based on video - Google Patents
A kind of water gauge reading detection method based on video Download PDFInfo
- Publication number
- CN109376740A CN109376740A CN201811222855.3A CN201811222855A CN109376740A CN 109376740 A CN109376740 A CN 109376740A CN 201811222855 A CN201811222855 A CN 201811222855A CN 109376740 A CN109376740 A CN 109376740A
- Authority
- CN
- China
- Prior art keywords
- water gauge
- water
- image
- digital
- gauge
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of, and the water gauge based on video reads detection method.This method obtains video image first, image is pre-processed, remove image noise, water gauge position is determined using the method for model inspection, and identification is split to the number on water gauge in water gauge region, sensed water level line position again, integrated digital segmentation recognition result and water level line testing result extrapolate water gauge reading, and this method is not easy camera subject angle, visual field variation influences and detects correct result.The present invention uses model inspection technology, can quickly position water gauge position;Using segmentation identification technology, it can effectively detect the position of water gauge number and digital reading is recognized accurately;Using gradient binary image floor projection technology, water level line position can be accurately positioned, to effectively increase the accuracy rate of water gauge reading.
Description
Technical field
The invention belongs to video detection technology field, reads detection side more particularly, to a kind of water gauge based on video
Method and device.
Background technique
In hydrological monitoring field, passes through eye-observation water gauge and determine water level reading, often expend a large amount of manpower
Cost and time cost, and read and can have large error, to cannot achieve the real-time accurate monitoring to water level reading.
Summary of the invention
The object of the present invention is to provide a kind of, and the water gauge based on video reads detection method, to be used for round-the-clock monitoring water gauge
Reading, and judge that water level reading whether within the scope of preset water level, achievees the effect that monitoring water level.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of water gauge reading detection method based on video, comprising:
(1) video image is pre-processed, removes image noise;
(2) using the position of water gauge in training pattern detection image, the region of water gauge is determined;
(3) in determining water gauge region to digital segmentation and identification;
(4) method for using binaryzation gradient image floor projection, determines water level line position;
(5) the digital segmentation recognition result in (3) and the water level line testing result in (4) are utilized, determines that current level is read
Number.
Further, in the step (1), image preprocessing mainly uses the method for gaussian filtering to the video figure of acquisition
As being filtered noise reduction process.
Further, it is specifically included in the step (2), acquires water gauge sample, training is based on the cascade sort of LBP feature
Device detects the water gauge in video image, then the water gauge that will test, as positive sample, erroneous detection is as negative sample, repeatedly
Training is iterated to model;Finally water gauge is detected with detection effect best sorter model, so that it is determined that water gauge
Specific location in the picture.
Further, in the step (3), using the water gauge region detected, according to the phase of number and water gauge on water gauge
To position, the specific location of water gauge numeric area in the picture is extrapolated, water gauge numeric area is subjected to binarization operation, is gone forward side by side
The horizontal exact position that each number on water gauge is determined with upright projection of row, then number is carried out using Gabor characteristic and neural network
Identification obtains specific location and digital recognition result digital on water gauge.
Another aspect of the present invention additionally provides a kind of water gauge reading detector based on video, comprising:
Preprocessing module: pre-processing video image, removes image noise;
Training detection module: using the position of water gauge in training pattern detection image, the region of water gauge is determined;
Divide identification module: in determining water gauge region to digital segmentation and identification;
Water level line detection module: with the method for binaryzation gradient image floor projection, water level line position is determined;
Water level reads module: utilizing the water in the digital segmentation recognition result and water level line detection module of segmentation identification module
Bit line testing result determines that current level is read.
Further, the preprocessing module includes gaussian filtering unit: with the method for gaussian filtering to the video of acquisition
Image is filtered noise reduction process.
Further, the trained detection module includes:
Classifier training unit: acquisition water gauge sample, training is based on LBP feature cascade classifier, in video image
Water gauge is detected;
Repetitive exercise unit: the water gauge that will test repeatedly changes to model as negative sample as positive sample, erroneous detection
Generation training;
Detection unit: water gauge is detected with detection effect best sorter model, so that it is determined that water gauge is in image
In specific location.
Further, the segmentation identification module includes:
Projected unit: water outlet is calculated according to the relative position of number and water gauge on water gauge using the water gauge region detected
The specific location of ruler numeric area in the picture;
Binarization unit: water gauge numeric area is subjected to binarization operation, and carry out level and upright projection determine water gauge
The exact position of upper each number;
Recognition unit: digital identification is carried out using Gabor characteristic and neural network, obtains specific location digital on water gauge
With digital recognition result.
Compared with prior art, the present invention have it is following the utility model has the advantages that
A kind of water gauge based on video provided by the invention reads detection method, by carrying out pretreatment to input picture
Except noise;Water gauge is detected using training pattern, determines water gauge position;Identification is split to number in water gauge region and determines water gauge
The position of number and reading;Floor projection, which is carried out, with the binary image of gradient image finds out water level line position;According to water level line
Digital segmentation recognition result on position and water gauge finds out the technical method of water gauge reading, realizes the automation of water gauge reading,
A large amount of human cost and time cost are not only eliminated, but also improves the accuracy and real-time of water gauge reading, the present invention
Applicability with several scenes.
Detailed description of the invention
Fig. 1 is the overall flow figure of the reading detection of water gauge described in the invention embodiment;
Fig. 2 is that number described in the invention embodiment identifies network structure.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
The present invention proposes that a kind of water gauge reading detection method based on video, overall flow figure are as shown in Figure 1.
1, video image is pre-processed, removes image noise.
Image is pre-processed using gaussian filtering in the present invention, can effectively inhibit noise, smoothed image.
2, using the position of water gauge in training pattern detection image, the region of water gauge is determined.
The cascade classifier based on LBP feature is used in the present invention, and water gauge is detected in entire image.LBP is special
Using centre of neighbourhood pixel as threshold value, the adjacent gray value of 8 pixels is compared with the pixel value of the centre of neighbourhood for the calculating of sign,
If surrounding pixel is greater than center pixel value, otherwise it is 0 that the position of the pixel, which is marked as 1,.The LBP of center pixel in this way
Value reflects the texture information of the pixel peripheral region.
The method that formula (1) center pixel LBP value calculates, wherein (xc, yc) be center pixel coordinate, p is the of neighborhood
P pixel, ipFor the gray value of neighborhood territory pixel, icFor the gray value of center pixel, s (x) is sign function.
The calculation method that LBP feature is calculated according to formula (1), is trained and is detected by the cascade classifier of SVM, root
The position of water gauge in the picture is determined according to testing result.
3, in determining water gauge region to digital segmentation and identification.
The number on water gauge is split using the method for both horizontally and vertically upright projection in the present invention, must be discharged
Digital position on ruler, then each number is known with the image-recognizing method based on Gabor direction character and neural network
Not.
The segmentation of water gauge number accounts for the ratio of water gauge according to number first, determines number column in the position in water gauge region
It sets, carries out binary conversion treatment with image of the binarization method based on textural characteristics to digital column region, then to binaryzation
It is both horizontally and vertically projected respectively, obtains both horizontally and vertically pixels statistics histogram, obtained often according to histogram
The right boundary and up-and-down boundary of a number, to obtain the position of number in the picture.
Number is identified using the image-recognizing method based on Gabor direction character and neural network.Utilize Gabor
Feature is extracted in transformation, if f (x, y) represents size as the two-dimentional water gauge Digital Character Image gray value of M × N, defines (xm, ym) be
Sampled point, then the Gabor characteristic of sampled point can indicate are as follows:
In formula (2), k=1,2,3...D-1, wavelength l value selects the directioin parameter θ between (3.2,3.8)kHave chosen 4
A direction value:This 4 values have respectively represented feature letters horizontal, vertical, on 4 directions of left and right diagonal line
Breath.
Neural metwork training is carried out for the Gabor characteristic on 4 directions extracted, network structure is as shown in Fig. 2, net
Network includes four sub-networks and an integrated network, and four sub-networks respectively identify the feature of four direction, overall network
Output for four sub-networks carries out comprehensive identification.Network is trained using BP algorithm, first trains 4 sub-networks,
Integrated network is trained in input using the output of 4 sub-networks as afterbody neural network again.
4, with the method for binaryzation gradient image floor projection, water level line position is determined.
Using the difference of background metope zone-texture where the water surface and water gauge in the present invention, the detection of water level line is carried out.It asks
The gradient image of entire image is taken, then by gradient image binary conversion treatment, floor projection then is carried out to binary image, is found out
Water level line position.
5, using digital segmentation recognition result and water level line testing result, determine that current level is read
Using position digital on the water gauge being partitioned into and the water level line position detected, and nearest apart from water level line
The recognition result of number, the comprehensive reading for extrapolating current level.
The foregoing is merely the preferred embodiments of the invention, are not intended to limit the invention creation, all at this
Within the spirit and principle of innovation and creation, any modification, equivalent replacement, improvement and so on should be included in the invention
Protection scope within.
Claims (8)
1. a kind of water gauge based on video reads detection method characterized by comprising
(1) video image is pre-processed, removes image noise;
(2) using the position of water gauge in training pattern detection image, the region of water gauge is determined;
(3) in determining water gauge region to digital segmentation and identification;
(4) method for using binaryzation gradient image floor projection, determines water level line position;
(5) the digital segmentation recognition result in (3) and the water level line testing result in (4) are utilized, determines that current level is read.
2. the method according to claim 1, wherein image preprocessing mainly uses Gauss in the step (1)
The method of filtering is filtered noise reduction process to the video image of acquisition.
3. water gauge sample is acquired the method according to claim 1, wherein specifically including in the step (2),
Training is based on LBP feature cascade classifier, detects to the water gauge in video image, then the water gauge that will test is as positive sample
This, erroneous detection is iterated training to model repeatedly as negative sample;Finally with the best sorter model of detection effect to water gauge
It is detected, so that it is determined that the specific location of water gauge in the picture.
4. the method according to claim 1, wherein in the step (3), using the water gauge region detected,
According to the relative position of number and water gauge on water gauge, the specific location of water gauge numeric area in the picture is extrapolated, by water gauge number
Block domain carries out binarization operation, and carry out level and upright projection determine the exact position of each number on water gauge, then use
Gabor characteristic and neural network carry out digital identification, obtain specific location and digital recognition result digital on water gauge.
5. a kind of water gauge reading detector based on video characterized by comprising
Preprocessing module: pre-processing video image, removes image noise;
Training detection module: using the position of water gauge in training pattern detection image, the region of water gauge is determined;
Divide identification module: in determining water gauge region to digital segmentation and identification;
Water level line detection module: with the method for binaryzation gradient image floor projection, water level line position is determined;
Water level reads module: utilizing the water level line in the digital segmentation recognition result and water level line detection module of segmentation identification module
Testing result determines that current level is read.
6. device according to claim 5, which is characterized in that the preprocessing module includes gaussian filtering unit: with height
The method of this filtering is filtered noise reduction process to the video image of acquisition.
7. device according to claim 5, which is characterized in that the trained detection module includes:
Classifier training unit: acquisition water gauge sample, training is based on LBP feature cascade classifier, to the water gauge in video image
It is detected;
Repetitive exercise unit: for the water gauge that will test as positive sample, erroneous detection is iterated instruction to model repeatedly as negative sample
Practice;
Detection unit: detecting water gauge with the best sorter model of detection effect, so that it is determined that water gauge is in the picture
Specific location.
8. device according to claim 5, which is characterized in that the segmentation identification module includes:
Projected unit: water gauge number is extrapolated according to the relative position of number and water gauge on water gauge using the water gauge region detected
The specific location of block domain in the picture;
Binarization unit: water gauge numeric area is subjected to binarization operation, and carry out level and upright projection determine on water gauge respectively
The exact position of number;
Recognition unit: digital identification is carried out using Gabor characteristic and neural network, obtains specific location and number digital on water gauge
Word recognition result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811222855.3A CN109376740A (en) | 2018-10-19 | 2018-10-19 | A kind of water gauge reading detection method based on video |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811222855.3A CN109376740A (en) | 2018-10-19 | 2018-10-19 | A kind of water gauge reading detection method based on video |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109376740A true CN109376740A (en) | 2019-02-22 |
Family
ID=65401038
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811222855.3A Pending CN109376740A (en) | 2018-10-19 | 2018-10-19 | A kind of water gauge reading detection method based on video |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109376740A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110287953A (en) * | 2019-05-20 | 2019-09-27 | 湖北九感科技有限公司 | Water level automatic identifying method and device |
CN110427933A (en) * | 2019-06-20 | 2019-11-08 | 浙江大学 | A kind of water gauge recognition methods based on deep learning |
CN111024181A (en) * | 2019-11-20 | 2020-04-17 | 武汉世纪水元科技股份有限公司 | Visual sensing water level measuring system aiming at complex environment |
CN111160210A (en) * | 2019-12-24 | 2020-05-15 | 天津天地伟业机器人技术有限公司 | Video-based water flow velocity detection method and system |
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 |
CN112362900A (en) * | 2020-10-30 | 2021-02-12 | 天地伟业技术有限公司 | Unattended water level flow velocity monitoring camera and monitoring method |
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 |
CN113822105A (en) * | 2020-07-07 | 2021-12-21 | 湖北亿立能科技股份有限公司 | Artificial intelligence water level monitoring system based on online two classifiers of SVM water scale |
CN113819971A (en) * | 2020-07-07 | 2021-12-21 | 湖北亿立能科技股份有限公司 | Artificial intelligence water level monitoring system based on water, scale and floater semantic segmentation |
CN115439861A (en) * | 2022-09-30 | 2022-12-06 | 北京中盛益华科技有限公司 | Water gauge recognition method based on OCR |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101957920A (en) * | 2010-09-08 | 2011-01-26 | 中国人民解放军国防科学技术大学 | Vehicle license plate searching method based on digital videos |
CN104680375A (en) * | 2015-02-28 | 2015-06-03 | 优化科技(苏州)有限公司 | Identification verifying system for living human body for electronic payment |
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 |
CN108256493A (en) * | 2018-01-26 | 2018-07-06 | 中国电子科技集团公司第三十八研究所 | A kind of traffic scene character identification system and recognition methods based on Vehicular video |
CN108364010A (en) * | 2018-03-08 | 2018-08-03 | 广东工业大学 | A kind of licence plate recognition method, device, equipment and computer readable storage medium |
-
2018
- 2018-10-19 CN CN201811222855.3A patent/CN109376740A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101957920A (en) * | 2010-09-08 | 2011-01-26 | 中国人民解放军国防科学技术大学 | Vehicle license plate searching method based on digital videos |
CN104680375A (en) * | 2015-02-28 | 2015-06-03 | 优化科技(苏州)有限公司 | Identification verifying system for living human body for electronic payment |
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 |
CN108256493A (en) * | 2018-01-26 | 2018-07-06 | 中国电子科技集团公司第三十八研究所 | A kind of traffic scene character identification system and recognition methods based on Vehicular video |
CN108364010A (en) * | 2018-03-08 | 2018-08-03 | 广东工业大学 | A kind of licence plate recognition method, device, equipment and computer readable storage medium |
Non-Patent Citations (1)
Title |
---|
金连文和覃剑钊: "基于Gabor方向特征及神经网络的车牌灰度字符图像识别", 《计算机工程》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110287953A (en) * | 2019-05-20 | 2019-09-27 | 湖北九感科技有限公司 | Water level automatic identifying method and device |
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 |
CN110427933A (en) * | 2019-06-20 | 2019-11-08 | 浙江大学 | A kind of water gauge recognition methods based on deep learning |
CN111024181A (en) * | 2019-11-20 | 2020-04-17 | 武汉世纪水元科技股份有限公司 | Visual sensing water level measuring system aiming at complex environment |
CN111160210A (en) * | 2019-12-24 | 2020-05-15 | 天津天地伟业机器人技术有限公司 | Video-based water flow velocity detection method and system |
CN111160210B (en) * | 2019-12-24 | 2023-09-26 | 天地伟业技术有限公司 | Video-based water flow rate detection method and system |
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 |
CN113822105A (en) * | 2020-07-07 | 2021-12-21 | 湖北亿立能科技股份有限公司 | Artificial intelligence water level monitoring system based on online two classifiers of SVM water scale |
CN113819971A (en) * | 2020-07-07 | 2021-12-21 | 湖北亿立能科技股份有限公司 | Artificial intelligence water level monitoring system based on water, scale and floater semantic segmentation |
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 |
CN113822105B (en) * | 2020-07-07 | 2024-04-19 | 湖北亿立能科技股份有限公司 | Artificial intelligence water level monitoring system based on online two classifiers of SVM water scale |
CN112362900A (en) * | 2020-10-30 | 2021-02-12 | 天地伟业技术有限公司 | Unattended water level flow velocity monitoring camera and monitoring method |
CN115439861A (en) * | 2022-09-30 | 2022-12-06 | 北京中盛益华科技有限公司 | Water gauge recognition method based on OCR |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109376740A (en) | A kind of water gauge reading detection method based on video | |
CN106557764B (en) | A kind of water level recognition methods based on binary-coded character water gauge and image procossing | |
Shi et al. | Automatic road crack detection using random structured forests | |
CN105046252B (en) | A kind of RMB prefix code recognition methods | |
CN107292310B (en) | Visual positioning and automatic reading method for circular pointer type dial plate | |
CN104200210B (en) | A kind of registration number character dividing method based on component | |
CN109255350B (en) | New energy license plate detection method based on video monitoring | |
CN110866430B (en) | License plate recognition method and device | |
CN102254159B (en) | Interpretation method for digital readout instrument | |
CN111539330B (en) | Transformer substation digital display instrument identification method based on double-SVM multi-classifier | |
CN108564814A (en) | A kind of parking position detection method and device based on image | |
CN109840483B (en) | Landslide crack detection and identification method and device | |
CN113240623B (en) | Pavement disease detection method and device | |
CN107727658A (en) | Pantograph crackle on-line measuring device and method based on image procossing | |
CN108009574A (en) | A kind of rail clip detection method | |
Ying et al. | Research on an automatic counting method for steel bars' image | |
CN105787912A (en) | Classification-based step type edge sub pixel localization method | |
CN115841633A (en) | Power tower and power line associated correction power tower and power line detection method | |
CN108961262A (en) | A kind of Bar code positioning method under complex scene | |
CN110580697B (en) | Video image processing method and system for measuring thickness of fetal nape transparency from ultrasonic video image | |
CN106548195A (en) | A kind of object detection method based on modified model HOG ULBP feature operators | |
CN104299005A (en) | Head detection method and system | |
US10115195B2 (en) | Method and apparatus for processing block to be processed of urine sediment image | |
CN116486212A (en) | Water gauge identification method, system and storage medium based on computer vision | |
CN115880683A (en) | Urban waterlogging ponding intelligent water level detection method based on deep learning |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190222 |