CN108764229B - Water gauge image automatic identification method based on computer vision technology - Google Patents

Water gauge image automatic identification method based on computer vision technology Download PDF

Info

Publication number
CN108764229B
CN108764229B CN201810535364.8A CN201810535364A CN108764229B CN 108764229 B CN108764229 B CN 108764229B CN 201810535364 A CN201810535364 A CN 201810535364A CN 108764229 B CN108764229 B CN 108764229B
Authority
CN
China
Prior art keywords
water gauge
character
water
image
characters
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
Application number
CN201810535364.8A
Other languages
Chinese (zh)
Other versions
CN108764229A (en
Inventor
贾西平
陈桂君
林正春
刘少鹏
林智勇
黄锦丽
张倩
廖秀秀
柏柯嘉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong Polytechnic Normal University
Original Assignee
Guangdong Polytechnic Normal University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangdong Polytechnic Normal University filed Critical Guangdong Polytechnic Normal University
Priority to CN201810535364.8A priority Critical patent/CN108764229B/en
Publication of CN108764229A publication Critical patent/CN108764229A/en
Application granted granted Critical
Publication of CN108764229B publication Critical patent/CN108764229B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • G06V10/23—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition based on positionally close patterns or neighbourhood relationships
    • 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
    • G06V10/48—Extraction of image or video features by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/146—Aligning or centring of the image pick-up or image-field
    • G06V30/1475—Inclination or skew detection or correction of characters or of image to be recognised
    • G06V30/1478—Inclination or skew detection or correction of characters or of image to be recognised of characters or characters lines

Abstract

The invention discloses a water gauge image automatic identification method based on a computer vision technology, which comprises the following steps: s1, water gauge character dictionary making: intercepting standard water gauge characters shot at different angles for deformation and normalization processing, and taking the standard water gauge characters as a water gauge character template; templates made under the conditions of different angles, deformation and the like of the same water gauge character form a template group of the water gauge character; template groups of different water gauge characters form a water gauge character dictionary; the method reduces the influence of illumination on the image color, effectively solves the problem that the water gauge target is difficult to segment under the complex background condition, and improves the accuracy of segmentation of the water gauge and the background; the character recognition rate is improved, and the effective recognition of characters can be realized for water gauge images with incomplete or abnormal characters; the water gauge reading rate of accuracy is high, and the maximum error of water level calculation result is within 4mm, satisfies actual water level monitoring demand.

Description

Water gauge image automatic identification method based on computer vision technology
Technical Field
The invention relates to the technical field of water level detection, in particular to a water gauge image automatic identification method based on a computer vision technology.
Background
The existing water gauge and background segmentation method generally adopts an edge-based segmentation algorithm, such as filtering, enhancing and other processing based on a gray level image; or the hough transformation is used for identifying the boundary of the water gauge, most of background and noise can be removed by the method, but the segmentation effect is easily influenced by factors such as illumination, water shoreline and the like, and the water gauge cannot be effectively segmented;
for the identification of the scale lines of the water gauge, the scale lines of the water gauge are extracted through projection analysis, K-means cluster analysis and the like to obtain the reading of the water gauge, the calculation process is simple and effective, but the complete water gauge image is not shot or the water gauge image with broken scale lines cannot be accurately identified. For the recognition of the water gauge characters, the existing character recognition method has classification and matching, and can solve the problem of recognition failure caused by local deletion or digital blurring of the water gauge characters to a certain extent. But the prepared samples are insufficient in abundance, the water gauge character images shot non-horizontally cannot be identified, and the character identification rate is low; for the water level calculation method, the mathematical relation among characters is mostly adopted, and the existing water level calculation result has errors of +/-1 cm and larger errors.
In conclusion, the existing water gauge image identification method has the problems that the water gauge image is not easy to segment due to the influence of complex backgrounds such as water lines and water shorelines, the water gauge character identification rate is not high due to insufficient sample richness, the water level calculation error is large and the like.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a water gauge image automatic identification method based on a computer vision technology, which can realize effective segmentation of a water gauge and a background under complex backgrounds such as water marks and the like, effective identification of water gauge characters under different degrees of inclination and high-precision water level value calculation.
The purpose of the invention is realized by the following technical scheme:
a water gauge image automatic identification method based on computer vision technology comprises the following steps:
s1, water gauge character dictionary making: intercepting standard water gauge characters shot at different angles for deformation and normalization processing, and taking the standard water gauge characters as a water gauge character template; templates made under different angles, deformation and other conditions of the same water gauge character form a template group of the character; template groups of different water gauge characters form a water gauge character dictionary;
s2, standing the water gauge in the water area to be detected, acquiring the water gauge image through the camera, inputting the water gauge image into the computer for the following processing:
s2.1, reading the image of the water ruler by the computer;
s2.2, background segmentation: firstly, down-sampling an acquired water gauge image, namely an RGB image, and converting the down-sampled image into an HSV image; then, opening the HSV image by using the circular structural elements, and extracting a water gauge outline and a background image; subtracting the extracted background from the HSV image; finally, the water gauge and the background are divided through an OSTU threshold, and noise with the area around the water gauge image smaller than a preset threshold is removed through area denoising;
s2.3, detecting a water gauge: detecting the edge of the water gauge by using a Canny operator, and performing linear detection on the water gauge edge image by combining hough transformation; rotating the water gauge binary image according to the detected inclination angles of the left edge and the right edge of the water gauge; projecting the water gauge image after inclination correction in the horizontal and vertical directions, and determining the positions of the upper edge, the lower edge, the left edge and the right edge of the water gauge according to the points of which the first value and the last value in the projected row pixel sum are both greater than zero and the points of which the first value and the last value in the column pixel sum are both greater than zero, and cutting the positions;
s2.4, dividing the water gauge characters: dividing the cut water gauge into two images with the same size on the left and the right according to the central line of the cut water gauge, horizontally projecting the left image, positioning the upper and the lower boundaries of the water gauge character, dividing the water gauge character according to the boundaries and removing redundant backgrounds; normalizing the size of the divided water gauge characters;
s2.5, waterAnd (3) identifying characters of the ruler: matching the water gauge characters segmented in the S2.4 with each water gauge character template group of the water gauge character dictionary manufactured in the S1 respectively to realize automatic identification of the water gauge characters; the numbers and the characters E are alternately arranged in the water ruler, the left numbers correspond to the characters on the right side
Figure BDA0001677684430000031
Correcting and checking the recognition result of the water gauge characters;
s2.6, calculating the water level: each water gauge character "E" or
Figure BDA0001677684430000032
The height of (A) is 5cm, wherein the height of one peak, namely the black horizontal bar, is 1cm, and the distance between one valley, namely the two black horizontal bars, is 1 cm; analyzing the actual situation of the water gauge characters in the water image, simultaneously performing horizontal projection, and calculating the accurate water gauge reading h according to the projected peak-valley ratio through the mathematical relationship between the charactersc(ii) a Calculating the water level h according to the reading of the water gauge3＝h1-h2＝h1-h+hcWherein h is the length of the water gauge, hcThe height of the water gauge in the water, i.e. the reading of the water gauge, h2The height of the water gauge exposed to the water surface, h1The height from the highest point of the water gauge to the water bottom;
s3, outputting water level value h by the computer3。
Preferably, in S2.5, the specific matching process is: according to the following formula (1),
Figure BDA0001677684430000033
the water gauge character to be recognized is the target character wqAnd any character w in the character template setiMatching is carried out, and the similarity Sim (w) between the two is calculatedq,wi) (ii) a Wherein m and n represent the number of rows and columns of the character image matrix respectively,
Figure BDA0001677684430000041
for water gauge characters w to be recognizedqThe average gray level of (a) is,
Figure BDA0001677684430000042
as a template character wi1,2,3, … …, 7;
according to the following formula (2),
Sim(wq,Sw)＝maxSim(wq,wi)，wi∈Sw (2)
with the character w to be recognizedqAnd character template set SwSimilarity of characters within Sim (w)q,wi) Maximum value of as wqAnd SwThe similarity of (c) is expressed as Sim (w)q,Sw)；
According to the following formula (3),
Figure BDA0001677684430000043
with the character w to be recognizedqSimilarity Sim (w) to each character template set in character dictionary Sq,Sw) Maximum value of as wqSimilarity to S, denoted Sim (w)q,S)；
According to the following formula (4),
Figure BDA0001677684430000044
outputting the character w to be recognizedqMatching result Q (w) with water gauge character dictionary SqS); wherein r is and wqThe most similar template character set represents a character, W is a water gauge character set, W is {0,1,2,3,4,5,6,7,8,9, E }, and Null represents WqAnd the matching is unsuccessful and cannot be identified.
Preferably, in S2.6, the water gauge reads hcThe calculation process of (2) is as follows:
if the character submerged by water on the left side of the water gauge is 'E', horizontally projecting the character image, and reading the water gauge according to the characteristic that the projected image is 'peak valley peak', wherein:
Figure BDA0001677684430000045
wherein h iscThe unit is cm, m is the value of the smallest numeric character in the recognition result, l1Height of the part of the character exposed to the water,/2Is the character "E" or
Figure BDA0001677684430000051
Width of the post-projection peak, l1And l2The ratio of (a) to (b) is the height (in cm) of the part of the character exposed to the water;
if the character submerged in water on the left side of the water gauge is a number, intercepting the character on the right side
Figure BDA0001677684430000052
Following the same procedure, the water gauge reads:
Figure BDA0001677684430000053
compared with the prior art, the invention has the following beneficial effects:
(1) the method reduces the influence of illumination on the image color, effectively solves the problem that the water gauge target is difficult to segment under the complex background condition, and improves the accuracy of segmentation of the water gauge and the background;
(2) the invention improves the character recognition rate, and can also realize the effective recognition of characters for the water gauge images with incomplete or abnormal characters;
(3) the water gauge has high reading accuracy, the maximum error of the water level calculation result is within 4mm, and the actual water level monitoring requirement is met.
Drawings
FIG. 1 is a general flow diagram of the present invention;
FIG. 2 is a background segmentation flow chart according to the present invention;
FIG. 3 is a flow chart of a multi-template matching character recognition algorithm of the present invention;
FIG. 4 is a schematic view of the water gauge reading calculation of the present invention;
fig. 5 is a schematic view of the water level measurement of the present invention.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
The basic principle of the invention is that a camera is used for acquiring a water gauge image, the acquired image is processed by a digital image processing technology, a computer is used for simulating human eyes to identify the reading of the water gauge, and the water level value is automatically read. The method is not influenced by external interference factors such as water quality, water temperature and sand content, and has strong applicability. The automatic acquisition of the water level comprises the following three aspects: the method comprises the following steps of effectively segmenting the water gauge from the background, effectively identifying the water gauge characters or scale marks and accurately calculating the water level.
As shown in fig. 1 to 5, a method for automatically identifying a water gauge image based on a computer vision technology includes the following steps:
(1) and (3) manufacturing a water gauge character dictionary: intercepting standard water gauge characters shot at different angles for deformation and normalization processing, and taking the standard water gauge characters as a water gauge character template; templates made under the conditions of different angles, deformation and the like of the same water gauge character form a template group of the water gauge character; template groups of different water gauge characters form a water gauge character dictionary;
(2) the water gauge is erected in a water area needing to be detected, a camera is used for acquiring a water gauge image, and the water gauge image is input into a computer for the following processing:
(2.1) reading the water ruler image by the computer;
(2.2) background segmentation: as shown in fig. 2, firstly, down-sampling an acquired water gauge image, that is, an RGB image, and converting the down-sampled image into an HSV image; then, opening the HSV image by using the circular structural elements, and extracting a water gauge outline and a background image; subtracting the extracted background from the HSV image; finally, the water gauge and the background are divided through an OSTU threshold, and noise with the area around the water gauge image smaller than a preset threshold is removed through area denoising;
in order to eliminate complex backgrounds such as water marks and water shorelines and effectively segment the water gauge and the background, down-sampling is carried out on an image, an RGB color space is converted into an HSV color space, the background is extracted by a morphological method, and the segmentation of the water gauge and the background is realized by the steps of enhancing, binarizing, denoising and the like.
(2.3) water gauge detection: detecting the edge of the water gauge by using a Canny operator, and performing linear detection on the water gauge edge image by combining hough transformation; rotating the water gauge binary image according to the detected inclination angles of the left edge and the right edge of the water gauge; projecting the water gauge image after inclination correction in the horizontal and vertical directions, and determining the positions of the upper edge, the lower edge, the left edge and the right edge of the water gauge according to the points of which the first value and the last value in the projected row pixel sum are both greater than zero and the points of which the first value and the last value in the column pixel sum are both greater than zero, and cutting the positions;
(2.4) dividing the water gauge characters: dividing the cut water gauge into two images with the same size on the left and the right according to the central line of the cut water gauge, horizontally projecting the left image, positioning the upper and the lower boundaries of the water gauge character, dividing the water gauge character according to the boundaries and removing redundant backgrounds; normalizing the size of the divided water gauge characters;
(2.5) water gauge character recognition: matching the water gauge characters segmented in the step (2.4) with each water gauge character template group of the water gauge character dictionary manufactured in the step (1) respectively to realize automatic identification of the water gauge characters; the numbers and the characters E are alternately arranged in the water ruler, the left numbers correspond to the characters on the right side
Figure BDA0001677684430000071
Correcting and checking the recognition result of the water gauge characters;
specifically, as shown in fig. 3, the multi-template matching water gauge character recognition algorithm: in order to improve the water gauge character recognition rate, standard water gauge characters shot at different angles are intercepted and subjected to normalization processing, and the standard water gauge characters are used as templates to enrich the water gauge character templates; template composition set of water gauge character w made under different conditionsSwThe template set of all the water gauge characters forms a set S, if S is availablewIs contained in S; the specific matching process is as follows: according to the following formula (1),
Figure BDA0001677684430000072
the water gauge character (i.e. target character) w to be recognizedqAnd any character w in the character template setiMatching is carried out, and the similarity Sim (w) between the two is calculatedq,wi) (ii) a Wherein m and n represent the number of rows and columns of the character image matrix respectively,
Figure BDA0001677684430000081
for water gauge characters w to be recognizedqThe average gray level of (a) is,
Figure BDA0001677684430000082
as a template character wi1,2,3, … …, 7;
according to the following formula (2),
Sim(wq,Sw)＝maxSim(wq,wi)，wi∈Sw (2)
with the character w to be recognizedqAnd character template set SwSimilarity of characters within Sim (w)q,wi) Maximum value of as wqAnd SwThe similarity of (c) is expressed as Sim (w)q,Sw)；
According to the following formula (3),
Figure BDA0001677684430000083
with the character w to be recognizedqSimilarity Sim (w) to each character template set in character dictionary Sq,Sw) Maximum value of as wqSimilarity to S, denoted Sim (w)q,S)。
According to the following formula (4),
Figure BDA0001677684430000084
outputting the character w to be recognizedqMatching result Q (w) with water gauge character dictionary SqS); wherein r is and wqThe most similar template character set represents a character, W is a water gauge character set, W is {0,1,2,3,4,5,6,7,8,9, E }, and Null represents WqAnd the matching is unsuccessful and cannot be identified.
(2.6) water level calculation: each water gauge character "E" or
Figure BDA0001677684430000085
The height of (A) is 5cm, wherein the height of one peak, namely the black horizontal bar, is 1cm, and the distance between one valley, namely the two black horizontal bars, is 1 cm; analyzing the actual situation of the water gauge characters in the water image, simultaneously performing horizontal projection, and calculating the accurate water gauge reading h according to the projected peak-valley ratio through the mathematical relationship between the charactersc；
Specifically, as shown in FIG. 4, the water level is calculated primarily by the water gauge reading hcObtaining, to obtain an accurate water gauge reading hcThe actual situation of the water gauge character in the water image is analyzed:
if the character submerged by water on the left side of the water gauge is 'E', horizontally projecting the character image, and reading the water gauge according to the characteristic that the projected image is 'peak valley peak', wherein:
Figure BDA0001677684430000091
wherein h iscThe unit is cm, m is the value of the smallest numeric character in the recognition result, l1Height of the part of the character exposed to the water,/2Is the character "E" or
Figure BDA0001677684430000092
Width of the post-projection peak, l1And l2Is the ratio ofThe height (in cm) of the part of the character exposed to the water surface;
if the character submerged in water on the left side of the water gauge is a number, intercepting the character on the right side
Figure BDA0001677684430000093
Following the same procedure, the water gauge reads:
Figure BDA0001677684430000094
as shown in FIG. 5, the water level h is calculated according to the water gauge reading3＝h1-h2＝h1-h+hcWherein h is the length of the water gauge, hcThe height of the water gauge in the water, i.e. the reading of the water gauge, h2The height of the water gauge exposed to the water surface, h1The height from the highest point of the water gauge to the water bottom and the highest water level of the basin in the past year can be obtained through the hydrological monitoring station;
(3) computer output water level value h3。
The method reduces the influence of illumination on the color of the image, effectively solves the problem that the target of the water gauge is difficult to segment under the complex background condition, and improves the accuracy of segmentation of the water gauge and the background, as shown in the following table 1, all evaluations of the segmentation method of the invention are higher than those of the prior art;
TABLE 1 segmentation method Performance comparison
Method Dice coefficient IoU Precision Recall
Document [2]]Method 0.770 0.637 0.708 0.896
Document [3]Method 0.823 0.700 0.702 0.995
The method of the invention 0.971 0.943 0.954 0.989
Note: document [2] chencui, liu zheng wei, cheng sheng, luomanna, niuxian, ruan. Document [3] high dawn, Wangzheng, Wangxin, Liuji Wei, HSV space-based video real-time water level detection algorithm [ J ]. Zhengzhou university proceedings (science edition), 2010,42(03): 75-79.
The invention improves the character recognition rate, and can also realize the effective recognition of characters for the water gauge images with incomplete or abnormal characters;
the water gauge has high reading accuracy, the maximum error of a water level calculation result is within 4mm, and the actual water level monitoring requirement is met as shown in the following table 2:
TABLE 2 Experimental data sheet
Serial number Inclination angle/° Conventional template matching recognition rate/%) Experimental recognition rate/%) Manual reading/cm Reading/cm of experiment Error/cm
1 -5.0 57.1 100 42.1 42.1 0
2 -3.7 71.4 100 48.5 48.9 +0.4
3 -7.8 75.0 100 49.0 48.9 -0.1
4 -1.5 83.3 100 37.2 37.2 0
5 -2.7 71.4 100 32.0 31.8 -0.2
6 -1.0 85.7 100 31.9 31.9 0
7 -4.3 62.5 100 37.8 37.8 0
8 -4.3 87.5 100 37.5 37.5 0
9 1.0 75.0 100 37.6 37.6 0
10 -1.0 71.4 100 31.9 31.9 0
Mean value - 74.0 100 - - -
The present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents and are included in the scope of the present invention.

Claims (3)

1. A water gauge image automatic identification method based on computer vision technology is characterized by comprising the following steps:
s1, water gauge character dictionary making: intercepting standard water gauge characters shot at different angles for deformation and normalization processing, and taking the standard water gauge characters as a water gauge character template; templates made under the conditions of different angles, deformation and the like of the same water gauge character form a template group of the water gauge character; template groups of different water gauge characters form a water gauge character dictionary;
s2, standing the water gauge in the water area to be detected, acquiring the water gauge image through the camera, inputting the water gauge image into the computer for the following processing:
s2.1, reading the image of the water ruler by the computer;
s2.2, background segmentation: firstly, down-sampling an acquired water gauge image, namely an RGB image, and converting the down-sampled image into an HSV image; then, opening the HSV image by using the circular structural elements, and extracting a water gauge outline and a background image; subtracting the extracted background from the HSV image; finally, the water gauge and the background are divided through an OSTU threshold, and noise with the area around the water gauge image smaller than a preset threshold is removed through area denoising;
s2.3, detecting a water gauge: detecting the edge of the water gauge by using a Canny operator, and performing linear detection on the water gauge edge image by combining hough transformation; rotating the water gauge binary image according to the detected inclination angles of the left edge and the right edge of the water gauge; projecting the water gauge image after inclination correction in the horizontal and vertical directions, and determining the positions of the upper edge, the lower edge, the left edge and the right edge of the water gauge according to the points of which the first value and the last value in the projected row pixel sum are both greater than zero and the points of which the first value and the last value in the column pixel sum are both greater than zero, and cutting the positions;
s2.4, dividing the water gauge characters: dividing the cut water gauge into two images with the same size on the left and the right according to the central line of the cut water gauge, horizontally projecting the left image, positioning the upper and the lower boundaries of the water gauge character, dividing the water gauge character according to the boundaries and removing redundant backgrounds; normalizing the size of the divided water gauge characters;
s2.5, water gauge character recognition: matching the water gauge characters segmented in the S2.4 with each water gauge character template group of the water gauge character dictionary manufactured in the S1 respectively to realize automatic identification of the water gauge characters; the numbers and the characters E are alternately arranged in the water ruler, the left numbers correspond to the characters on the right side
Figure FDA0001677684420000024
Correcting and checking the recognition result of the water gauge characters;
s2.6, calculating the water level: each water gauge character "E" or
Figure FDA0001677684420000025
The height of (A) is 5cm, wherein the height of one peak, namely the black horizontal bar, is 1cm, and the distance between one valley, namely the two black horizontal bars, is 1 cm; analyzing the actual situation of the water gauge characters in the water image, simultaneously performing horizontal projection, and calculating the accurate water gauge reading h according to the projected peak-valley ratio through the mathematical relationship between the charactersc(ii) a Calculating the water level h according to the reading of the water gauge3＝h1-h2＝h1-h+hcWherein h is the length of the water gauge, hcThe height of the water gauge in the water, i.e. the reading of the water gauge, h2The height of the water gauge exposed to the water surface, h1The height from the highest point of the water gauge to the water bottom;
s3, outputting water level value h by the computer3。
2. The method for automatically identifying the water gauge image based on the computer vision technology according to the claim 1, wherein in the S2.5, the specific matching process is as follows: according to the following formula (1),
Figure FDA0001677684420000021
the water gauge character to be recognized is the target character wqAnd any character w in the character template setiTo carry outMatching, calculating the similarity Sim (w) between the twoq,wi) (ii) a Wherein m and n represent the number of rows and columns of the character image matrix respectively,
Figure FDA0001677684420000022
for water gauge characters w to be recognizedqThe average gray level of (a) is,
Figure FDA0001677684420000023
as a template character wi1,2,3, … …, 7;
according to the following formula (2),
Sim(wq,Sw)＝maxSim(wq,wi)，wi∈Sw (2)
with the character w to be recognizedqAnd character template set SwSimilarity of characters within Sim (w)q,wi) Maximum value of as wqAnd SwThe similarity of (c) is expressed as Sim (w)q,Sw)；
According to the following formula (3),
Figure FDA0001677684420000033
with the character w to be recognizedqSimilarity Sim (w) with each character template set in water gauge character dictionary Sq,Sw) Maximum value of as wqSimilarity to S, denoted Sim (w)q,S)；
According to the following formula (4),
Figure FDA0001677684420000031
outputting the character w to be recognizedqMatching result Q (w) with water gauge character dictionary SqS); wherein r is and wqThe most similar template character set represents a character, W is a water gauge character set, W is {0,1,2,3,4,5,6,7,8,9, E }, and Null represents WqIs not matched intoAnd successfully, the identification cannot be carried out.
3. The method for automatically identifying the water gauge image based on the computer vision technology as claimed in claim 1, wherein in S2.6, the water gauge reading hcThe calculation process of (2) is as follows:
if the character submerged by water on the left side of the water gauge is 'E', horizontally projecting the character image, and reading the water gauge according to the characteristic that the projected image is 'peak valley peak', wherein:
Figure FDA0001677684420000032
wherein h iscThe unit is cm, m is the value of the smallest numeric character in the recognition result, l1Height of the part of the character exposed to the water,/2Is the character "E" or
Figure FDA0001677684420000034
Width of the post-projection peak, l1And l2The ratio of (a) to (b) is the height (in cm) of the part of the character exposed to the water;
if the character submerged in water on the left side of the water gauge is a number, intercepting the character on the right side
Figure FDA0001677684420000042
Following the same procedure, the water gauge reads:
Figure FDA0001677684420000041
CN201810535364.8A 2018-05-29 2018-05-29 Water gauge image automatic identification method based on computer vision technology Active CN108764229B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810535364.8A CN108764229B (en) 2018-05-29 2018-05-29 Water gauge image automatic identification method based on computer vision technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810535364.8A CN108764229B (en) 2018-05-29 2018-05-29 Water gauge image automatic identification method based on computer vision technology

Publications (2)

Publication Number Publication Date
CN108764229A CN108764229A (en) 2018-11-06
CN108764229B true CN108764229B (en) 2021-05-18

Family

ID=64003861

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810535364.8A Active CN108764229B (en) 2018-05-29 2018-05-29 Water gauge image automatic identification method based on computer vision technology

Country Status (1)

Country Link
CN (1) CN108764229B (en)

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109711353B (en) * 2018-12-28 2023-06-30 煤炭科学研究总院有限公司 Ship waterline area identification method based on machine vision
CN109766886B (en) * 2019-01-02 2023-04-28 南京林业大学 Water level identification method based on image identification
CN110031060B (en) * 2019-05-14 2020-11-10 武汉大学 Method for measuring precision of projection amplification water level and application
US11443133B2 (en) 2019-12-10 2022-09-13 Ricoh Company, Ltd. Computer vision system for industrial equipment gauge digitization and alarms
CN111652089B (en) * 2020-05-18 2023-09-19 纪义胜 Automatic water level identification method and system based on image processing
CN111626190B (en) * 2020-05-26 2023-07-07 浙江大学 Water level monitoring method for scale recognition based on clustering partition
CN111998910B (en) * 2020-08-26 2021-09-24 河海大学 Visual measurement method and system for water level of multi-stage water gauge
CN112329644A (en) * 2020-11-06 2021-02-05 中冶赛迪重庆信息技术有限公司 Reservoir water level monitoring method and system, medium and electronic terminal
CN113221898B (en) * 2021-04-16 2022-02-15 北京科技大学 Automatic water level gauge reading method
CN113392846A (en) * 2021-06-16 2021-09-14 广州观必达数据技术有限责任公司 Water gauge water level monitoring method and system based on deep learning
CN115578695B (en) * 2022-11-21 2023-03-28 昆明理工大学 Water gauge water level machine vision detection method and device with free shooting visual angle
CN115546235B (en) * 2022-11-23 2023-04-21 成都见海科技有限公司 Water level identification method, system and storage medium based on image segmentation
CN116109869B (en) * 2023-02-20 2024-02-20 北京科技大学 Method and device for identifying reservoir water level fluctuation based on water bank image matching
CN116222693B (en) * 2023-05-10 2023-07-25 安徽省（水利部淮河水利委员会）水利科学研究院（安徽省水利工程质量检测中心站） Natural river channel water level automatic monitoring method based on water gauge reflection

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103487033A (en) * 2013-09-22 2014-01-01 河海大学 River surface photographic surveying method based on height-change homography
KR20170086773A (en) * 2016-01-19 2017-07-27 이현석 Apparatus for visual water level assessment

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070242884A1 (en) * 2001-07-24 2007-10-18 Dugan John P Method for measuring water depths using visible images of shoaling ocean waves

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103487033A (en) * 2013-09-22 2014-01-01 河海大学 River surface photographic surveying method based on height-change homography
KR20170086773A (en) * 2016-01-19 2017-07-27 이현석 Apparatus for visual water level assessment

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
《Applications of Image Recognition for Real-Time Water Level and Surface Velocity》;Franco Lin等;《 2013 IEEE International Symposium on Multimedia》;20140224;全文 *
《基于图像处理的水位信息自动提取技术》;陈翠 等;《水利信息化》;20160229;全文 *

Also Published As

Publication number Publication date
CN108764229A (en) 2018-11-06

Similar Documents

Publication Publication Date Title
CN108764229B (en) Water gauge image automatic identification method based on computer vision technology
CN111626190B (en) Water level monitoring method for scale recognition based on clustering partition
CN106960208B (en) Method and system for automatically segmenting and identifying instrument liquid crystal number
CN107633192B (en) Bar code segmentation and reading method based on machine vision under complex background
CN105067638B (en) Tire fetal membrane face character defect inspection method based on machine vision
CN112598001B (en) Automatic ship water gauge reading identification method based on multi-model fusion
CN110120042B (en) Crop image pest and disease damage area extraction method based on SLIC super-pixel and automatic threshold segmentation
CN110286126A (en) A kind of wafer surface defects subregion area detecting method of view-based access control model image
CN108960237B (en) Reading identification method for pointer type oil level indicator
CN109886954B (en) Printed matter defect detection method
CN111879241A (en) Mobile phone battery size measuring method based on machine vision
CN112734729B (en) Water gauge water level line image detection method and device suitable for night light supplement condition and storage medium
CN112651968A (en) Wood board deformation and pit detection method based on depth information
CN111598856A (en) Chip surface defect automatic detection method and system based on defect-oriented multi-point positioning neural network
CN109359604B (en) Method for identifying instrument under shadow interference facing inspection robot
CN110648330A (en) Defect detection method for camera glass
CN113870202A (en) Far-end chip defect detection system based on deep learning technology
CN113705564B (en) Pointer type instrument identification reading method
CN112419225B (en) SOP type chip detection method and system based on pin segmentation
CN112364862B (en) Histogram similarity-based disturbance deformation Chinese character picture matching method
CN111539330B (en) Transformer substation digital display instrument identification method based on double-SVM multi-classifier
CN114627463A (en) Non-contact power distribution data identification method based on machine identification
CN113902894A (en) Strip type level meter automatic reading identification method based on image processing
CN111339937B (en) Method for extracting handwritten data in key column in simple table photo
CN113159027A (en) Seven-segment type digital display instrument identification method based on minimum external rectangle variant

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