CN109784257A - A kind of detection of transformer thermometer and recognition methods - Google Patents
A kind of detection of transformer thermometer and recognition methods Download PDFInfo
- Publication number
- CN109784257A CN109784257A CN201910014109.3A CN201910014109A CN109784257A CN 109784257 A CN109784257 A CN 109784257A CN 201910014109 A CN201910014109 A CN 201910014109A CN 109784257 A CN109784257 A CN 109784257A
- Authority
- CN
- China
- Prior art keywords
- pointer
- image
- transformer
- candidate area
- transformer thermometer
- 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
Abstract
The invention discloses a kind of detection of transformer thermometer and recognition methods, and the method includes firstly, crusing robot, which reaches designated position, obtains picture；Secondly, coarse positioning and accurate positioning are carried out to target transformer temperature table section to be detected, it calculates perceptual hash index, mutual information index and the friendship of the multiple object candidate area and compares, and the multiple object candidate area is screened, obtain final goal image；Then, target image is read in, different image pretreatment operations is used according to color of pointer；Then, instrument is positioned with object detector, is partitioned into dial plate region, exclude the disturbing factor outside transformer thermometer dial plate；Then, pointer profile is extracted using correlation method according to color of pointer；Finally, calculating the reading of pointer direction of rotation and direction using the cosine law；The transformer thermometer identified under a variety of illumination, attitudes vibration can be detected using this method.
Description
Technical field
The present invention relates to electric inspection process robot fields, more particularly to a kind of detection of transformer thermometer and identification side
Method.
Background technique
Electric inspection process robot needs to realize autonomous localization and navigation in substation, the identification of field instrument registration, fills automatically
The basic functions such as electricity.Wherein the function of core is exactly the registration of the instrument and meter of detection and the live power equipment of identification.
Currently, electric inspection process robot is during detecting and identifying transformer thermometer registration, there are biggish difficulties
Degree.Most of transformer thermometer does not have the transporting function of intelligence instrument due to cost and history, can only patrol
Inspection robot goes to read instrument registration by the method for computer vision.And the premise for accurately identifying transformer thermometer registration is
Accurately detect the position of transformer thermometer in visual pattern, and most of transformer thermometers are in outdoor, it is most of at present
Method is detected and is identified using traditional image processing means, and in the case where illumination condition variation, detection effect is not
Good, generally a kind of illumination condition just needs one group of parameter, and in identification process, there is also outdoor different illumination, crusing robot
Situations such as different shooting angles.This just needs to propose a kind of more general detection and recognition methods, copes with different illumination, posture
Under the conditions of transformer thermometer detection and identification mission.
Summary of the invention
The purpose of the present invention is to provide a kind of detection of transformer thermometer and recognition methods, for live indicating transformer
Registration identification problem of the thermometer under the working conditions such as different illumination, posture optimizes.
The technical solution of the object of the invention are as follows: a kind of detection of transformer thermometer and recognition methods, the method packet
It includes,
Step 1: crusing robot reaches designated position and obtains transformer thermometer picture to be detected；
Step 2: it is based on the transformer thermometer picture to be detected, target transformer thermometer is positioned:
Coarse positioning is carried out to the target transformer temperature table section using plum forests Fourier transformation and phase coherent techniques,
Obtain coarse positioning target transformer temperature table section；
The target transformer temperature table section is accurately positioned using the method for machine learning, it will be described to be detected
Transformer thermometer picture is sent into classifier, obtains multiple object candidate areas；
It calculates perceptual hash index, mutual information index and the friendship of the multiple object candidate area and compares, and to described more
A object candidate area is screened, and final goal image is obtained；
Step 3: reading in the final goal image, pretreatment operation is carried out to the final goal image；
Step 4: positioning instrument using object detector, be partitioned into the table of transformer thermometer in the final goal image
Disk area；
Step 5: according to color of pointer, extracting pointer profile；
Step 6: calculating the reading of pointer direction of rotation and direction.
Further, further include in the step 1,
One is chosen in the inspection point using transformer thermometer image data set training classifier, and for each inspection point
The transformer thermometer of shooting image placed in the middle is as template image.
Further, the friendship of the multiple object candidate area is calculated in the step 2 and ratio, mutual information index and perception are breathed out
Uncommon index, and the multiple object candidate area is screened, obtaining final goal image includes:
The friendship of each object candidate area and the coarse positioning target transformer temperature table section and ratio are sought respectively；
Calculate separately the mutual information index of each object candidate area image and template image；
The perception of the transformer thermometer area image in each object candidate area image and template image is sought respectively
Hash index；
By the friendship of each object candidate area and three kinds of ratio, mutual information index and perceptual hash index indexs are done
Ranking operation finds out the confidence level of each object candidate area, using the maximum object candidate area of confidence level as standby
Select testing result；
If the satisfaction of alternative testing result is less than given threshold thresholdIOU, and (pHash+1/I (G(X),H(y))) big
When threshold value thresholdA, using the coarse positioning target transformer temperature table section as final goal image；Otherwise, with institute
Alternative testing result is stated as final goal image；Wherein, pHash is alternative testing result perceptual hash index, I (G(X),
H(Y)) it is mutual information index.
Further, the calculation formula of the confidence level are as follows:
Confidence=1- (pHash+1/I (G(X),H(y)))/(IOU+D)
In formula, I (G(X),H(Y)) be the target candidate area mutual information index, pHash be the target candidate area sense
Know Hash index, IOU is the friendship of object candidate area and coarse positioning target transformer temperature table section and than index, and D is setting
Constant.
Further, the threshold value thresholdIOU value range 0.1~0.4, the threshold value thresholdA value model
Enclose 10~50.
Further, include: to the final goal image pretreatment operation in the step 3
If pointer is red, HSV format is converted by the final goal image；
If pointer is non-red, by final goal image gray processing processing.
Further, pointer profile is extracted in the step 5 includes,
If pointer is red, specifically includes the following steps:
Red area is extracted using the color continuity of HSV, H value range is (0,10) (156,180), channel S value
Range is (43,255), and the channel V value range is (46,255)；
Carry out closed operation；
Extract pointer profile；
If the non-red of pointer, specifically includes the following steps:
Histogram equalization, gaussian filtering are carried out to image；
Using Otsu algorithm to dial plate region binaryzation, black pointer is become into white area, other backgrounds become black；
Opening operation is carried out, pointer profile is extracted.
Further, the step 6 includes,
Extract the profile point set of the pointer profile；
It finds in the profile meeting point at a distance of maximum two points, straight line is done with this two o'clock；
Continue to find the point pair on pointer profile, form it into the line segment vertical with the straight line, and from described centering
Find the point pair of lie farthest away；
The point of the lie farthest away is the circle of the transformer thermometer to the line segment of formation and the point of the straight line intersection
The heart establishes coordinate system by origin of the center of circle；
Based on any two in vertical axis in the vector and coordinate system that apart maximum two points are formed in profile meeting point
Point vector, the reading of pointer direction of rotation and direction is calculated using the cosine law.
Compared with prior art, the present invention its remarkable advantage are as follows: (1) present invention has merged the existing positioning letter of robot
Breath, due to the accuracy of robot navigation's location technology, position error is less than 5cm, solves robot location's not timing, target
The big problem of scale, angle change；(2) variations such as graphical rule, the rotation that camera is taken are smaller, utilize phase on this basis
Position is related to carry out rough detection, for image illumination variation, detect the attitudes vibration of target, carried out using the method for machine learning
The problems such as accurate detection solves target and is illuminated by the light influence greatly, such as excessively bright, excessively dark；(3) after detecting target image, will refer to
Needle respectively pre-processes image with non-red by red, and by closed operation, the operations such as histogram equalization improve transformation
Device thermometer registration recognition accuracy preferably solves the transformer thermometer registration identification under the conditions of different illumination, posture
Problem improves robot routing inspection efficiency.(4) hsv color space is utilized, the influence that illumination identifies registration is reduced.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
It obtains it is clear that understand through the implementation of the invention.The objectives and other advantages of the invention can be by specification, right
Pointed structure is achieved and obtained in claim and attached drawing.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair
Bright some embodiments for those of ordinary skill in the art without creative efforts, can be with root
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is transformer thermometer detection a kind of in the embodiment of the present invention and recognition methods flow diagram；
Fig. 2 is the detailed process of transformer thermometer detection and recognition methods kind step 3- step 5 in the embodiment of the present invention
Schematic diagram；
Fig. 3 is perceptual hash calculating process schematic diagram in the embodiment of the present invention；
Fig. 4 is that pointer direction of rotation and the direction reading cosine law calculate schematic diagram in the embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical solution in the embodiment of the present invention clearly and completely illustrated, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
As shown in Figure 1, describing a kind of detection of transformer thermometer and recognition methods in the embodiment of the present invention, including following
Step:
Step 1: crusing robot reaches designated position and obtains transformer thermometer picture to be detected；
Step 2: being based on the transformer thermometer picture to be detected, transformer temperature table section to be detected is carried out thick
Positioning and accurate positioning, screening object candidate area obtain final goal image；
Step 3: reading in the final goal image, different image pretreatment operations is used according to color of pointer；
Step 4: the target transformer temperature table section is positioned with object detector, so that meter location is in picture
Centre, is then partitioned into dial plate region；To exclude the disturbing factor outside transformer thermometer dial plate using this method；Then direct root
The middle section of picture after pretreatment is partitioned into according to coordinate, this region includes dial plate, consequently facilitating pointer extracting.
Step 5: pointer profile is extracted according to presetting method according to color of pointer；
Step 6: the reading of pointer direction of rotation and direction is calculated using the cosine law；
Further, step 1 concrete operations are as follows: utilize the Instrument image data set training classification of transformer thermometer
Device, and a transformer thermometer image placed in the middle in inspection point shooting is chosen as template image for each inspection point,
Crusing robot reaches specified inspection point and obtains instrument picture to be detected.Preferably, the classifier is Adaboost classifier.
Further, the concrete operations of the step 2 are as follows:
Step 2.1: using plum forests Fourier transformation and phase coherent techniques in transformer thermometer picture to be detected
Target transformer temperature table section carries out coarse positioning；
Step 2.2: target transformer temperature table section is accurately positioned using the method for machine learning, it will be to be detected
Transformer thermometer picture is sent into trained listening group, obtains several object candidate areas；
Step 2.3: calculating perceptual hash, mutual information and the friendship of multiple candidate regions and three kinds of parameter indexes of ratio, screen mesh
Mark candidate region obtains final goal.
Further, step 2.3 the following steps are included:
Step 2.3.1: respectively by each object candidate area and the coarse positioning target transformer temperature table section
Ask friendship and than parameter IOU；
Step 2.3.2: respectively by the transformer thermometer administrative division map in each object candidate area image and template image
As doing perceptual hash calculating, perceptual hash index is obtained；
Step 2.3.3: the mutual information index of each the object candidate area image and template image is calculated separately；
Step 2.3.4: refer to by the friendship of each object candidate area and than IOU, mutual information index and perceptual hash
It marks three kinds of indexs and does the confidence level that weighting finds out each object candidate area, the maximum object candidate area of confidence level is made
For alternative testing result；
Step 2.3.5: if the IOU of alternative testing result meets while being less than given threshold thresholdIOU, and
(pHash+1/I(G(X),H(y))) be greater than threshold value thresholdA when, will in step 2 determine the coarse positioning target transformer
Temperature table section is as final goal image, and otherwise in case selecting testing result as final goal image, pHash is alternative detection
As a result perceptual hash index, I (G(X),H(Y)) it is mutual information index.
Preferably, each object candidate area and the coarse positioning target instrument region are asked into friendship in step 2.3.1
And the formula than parameter IOU are as follows:
In formula, C is coarse positioning target instrument region, niFor i-th of object candidate area.
Preferably, respectively by the gauge field in each object candidate area image and template image in step 2.3.2
Area image does perceptual hash calculating method particularly includes:
A1, pretreatment: downscaled images size, and by image gray processing；
A2, dct transform: dct transform is carried out to pretreated image, obtains matrix F (u, v)；
A3, DCT matrix is reduced, as shown in figure 3, the feature of entire image concentrates on upper left corner low frequency region, we are extracted
Matrix top left corner pixel 8*8 matrix, the eigenmatrix as the image；
A4, matrix binaryzation in Fig. 3, average to matrix, and will be greater than mean value sets 1, sets 0 less than mean value；
A5, generate cryptographic Hash: the sequence for being 64 by the matrix arrangement after binaryzation, which is the Kazakhstan of input picture
Uncommon sequence；
The Hamming distance of the feature vector of A6, calculating object candidate area image and template image, perceptually Hash refers to
Mark；
Preferably, in step 2.3.3 each the object candidate area image and the template image mutual information index
Calculation formula are as follows:
G(X)、H(Y)The respectively number of template image and candidate image gray-scale pixels, W, H are respectively candidate region image
It is wide, high.
Preferably, in step 2.3.4 confidence level calculation formula are as follows:
Confidence=1- (pHash+1/I (G(X),H(y)))/(IOU+D)
In formula, I (G(X),H(Y)) it is mutual information index, pHash is perceptual hash index, and for IOU to hand over and than index, D is to set
Fixed constant.
Preferably, threshold value thresholdIOU value range 0.1~0.4, threshold value thresholdA value range 10~50.
Further, as shown in Fig. 2, describing the idiographic flow schematic diagram of step 3- step 5, wherein in the step 3
Robot shoots to specified inspection point and obtains transformer thermometer picture, and pointer is divided into red and non-red two types,
Corresponding different image processing method.Because picture is placed under hsv color space, black is vulnerable to illumination effect, in HSV sky
Between it is lower be easy it is fuzzy, and it is red illumination effect is not easily susceptible under HSV space, therefore for red pointer, picture is gone into HSV
Format is handled picture gray processing for non-red pointer.
Further, in Fig. 2, pointer is extracted using correlation method according to color of pointer.Under hsv color space, such as table 1
Shown, distribution of color is continuous, therefore according to following table, is red feature using color of pointer, extracts pointer part.For
Non- red pointer, picture gray processing is handled.
Table 1, the color value range under hsv color space
Further, step 5 concrete operations are as follows:
For red pointer:
Step 5.1.1: pointer is red, since transformer thermometer other parts do not have red area, utilizes the face of HSV
Color continuity extracts red area, H value in (0,10) (156,180), channel S value in (43,255), the channel V (46,
255)；
Step 5.1.2: carrying out closed operation, reduces noise point to the interference for extracting pointer and makes profile round and smooth.
Step 5.1.3: area maximum red area, as pointer profile are extracted.
For non-red pointer:
Step 5.2.1: carrying out histogram equalization, gaussian filtering to image, reduces illumination interference.
Step 5.2.2: since transformer temperature list index and dial plate color difference are obvious, therefore utilize Otsu algorithm to dial plate
Black pointer is become white area by region binaryzation, other backgrounds become black.
Step 5.2.3: carrying out opening operation, extracts pointer profile.
Further, as shown in figure 4, step 6 concrete operations are as follows:
Step 6.1: finding pointer area obtained in step 5 and extract the profile point set in the region, find out the set
In at a distance of remote two points most, straight line a is done with this two o'clock, looks for the point pair on profile, the straight line b formed it into perpendicular to straight line a,
The point pair of lie farthest away is found in from these.
Step 6.2: the point of straight line a, b intersection, the as center of circle O of transformer thermometer establish coordinate system by origin of O.
Step 6.3: the shape due to pointer integral into triangle, the vertex A and center of circle O lie farthest away of pointer, according to this
It is vertex A that feature, which finds point farthest in profile point set, is taken up an official post intention amount BC using vector AO and vertical axis, passes through cosine
Theorem calculates the reading of pointer direction of rotation and direction:
AOBC=| | AO | | | | BC | | cos θ
Θ=antcos (AOBC/ | | AO | | * | | BC | |)
Wherein AO, BC are vector, and AOBC is the inner product of vector, and θ is the angle of two vectors.
Compared with prior art, the present invention its remarkable advantage are as follows: (1) present invention has merged the existing positioning letter of robot
Breath.Due to the accuracy of robot navigation's location technology, position error is less than 5cm, solves robot location's not timing, target
The big problem of scale, angle change；(2) variations such as graphical rule, the rotation that camera is taken are smaller, utilize phase on this basis
Position is related to carry out rough detection, for image illumination variation, detect the attitudes vibration of target, carried out using the method for machine learning
The problems such as accurate detection solves target and is illuminated by the light influence greatly, such as excessively bright, excessively dark；(3) after detecting target image, will refer to
Needle respectively pre-processes image with non-red by red, and by closed operation, the operations such as histogram equalization improve transformation
Device thermometer registration recognition accuracy preferably solves the transformer thermometer registration identification under the conditions of different illumination, posture
Problem improves robot routing inspection efficiency.(4) hsv color space is utilized, the influence that illumination identifies registration is reduced.
Although the present invention is described in detail referring to the foregoing embodiments, those skilled in the art should manage
Solution: it is still possible to modify the technical solutions described in the foregoing embodiments, or to part of technical characteristic into
Row equivalent replacement；And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The spirit and scope of scheme.
Claims (8)
1. a kind of transformer thermometer detection and recognition methods, which is characterized in that the method includes,
Step 1: crusing robot reaches designated position and obtains transformer thermometer picture to be detected；
Step 2: being based on the transformer thermometer picture to be detected, screen the final goal figure of target transformer temperature table section
Picture:
Coarse positioning is carried out to the target transformer temperature table section using plum forests Fourier transformation and phase coherent techniques, is obtained
Coarse positioning target transformer temperature table section；
The target transformer temperature table section is accurately positioned using the method for machine learning, by the transformation to be detected
Device thermometer picture is sent into classifier, obtains multiple object candidate areas；
It calculates perceptual hash index, mutual information index and the friendship of the multiple object candidate area and compares, and to the multiple mesh
Mark candidate region is screened, and final goal image is obtained；
Step 3: reading in the final goal image, pretreatment operation is carried out to the final goal image；
Step 4: positioning instrument using object detector, be partitioned into the dial plate area of transformer thermometer in the final goal image
Domain；
Step 5: according to color of pointer, extracting pointer profile；
Step 6: calculating the reading of pointer direction of rotation and direction.
2. transformer thermometer detection according to claim 1 and recognition methods, which is characterized in that in the step 1 also
Including,
Using transformer thermometer image data set training classifier, and one is chosen for each inspection point and is shot in the inspection point
Transformer thermometer image placed in the middle as template image.
3. transformer thermometer detection according to claim 1 and recognition methods, which is characterized in that the step 2 is fallen into a trap
The friendship of the multiple object candidate area and ratio, mutual information index and perceptual hash index are calculated, and to the multiple target candidate
Region is screened, and is obtained final goal image and is included:
The friendship of each object candidate area and the coarse positioning target transformer temperature table section and ratio are sought respectively；
Calculate separately the mutual information index of each object candidate area image and template image；
The perceptual hash of the transformer thermometer area image in each object candidate area image and template image is sought respectively
Index；
By the friendship of each object candidate area and three kinds of ratio, mutual information index and perceptual hash index indexs weight
Operation finds out the confidence level of each object candidate area, and the maximum object candidate area of confidence level is alternately examined
Survey result；
If the friendship of alternative testing result is simultaneously less than given threshold thresholdIOU than meeting, and (pHash+1/I (G(X),H(y)))
When greater than threshold value thresholdA, using the coarse positioning target transformer temperature table section as final goal image；Otherwise, with
The alternative testing result is as final goal image；Wherein, pHash is alternative testing result perceptual hash index, I (G(X),
H(Y)) be the alternative testing result mutual information index.
4. transformer thermometer detection according to claim 3 and recognition methods, which is characterized in that the meter of the confidence level
Calculate formula are as follows:
Confidence=1- (pHash+1/I (G(X),H(y)))/(IOU+D)
In formula, I (G(X),H(Y)) be the target candidate area mutual information index, pHash be the target candidate area perception breathe out
Uncommon index, IOU are the friendship of object candidate area and coarse positioning target transformer temperature table section and than index, D be set it is normal
Number.
5. transformer thermometer detection according to claim 3 and recognition methods, which is characterized in that the threshold value
ThresholdIOU value range 0.1~0.4, the threshold value thresholdA value range 10~50.
6. transformer thermometer detection according to claim 1 and recognition methods, which is characterized in that right in the step 3
The final goal image pretreatment operation includes:
If pointer is red, HSV format is converted by the final goal image；
If pointer is non-red, by final goal image gray processing processing.
7. transformer thermometer detection according to claim 1 and recognition methods, which is characterized in that mentioned in the step 5
Fetching pinwheel exterior feature includes,
If pointer is red, specifically includes the following steps:
Red area is extracted using the color continuity of HSV, H value range is (0,10) (156,180), channel S value range
For (43,255), the channel V value range is (46,255)；
Carry out closed operation；
Extract pointer profile；
If the non-red of pointer, specifically includes the following steps:
Histogram equalization, gaussian filtering are carried out to image；
Using Otsu algorithm to dial plate region binaryzation, black pointer is become into white area, other backgrounds become black；
Opening operation is carried out, pointer profile is extracted.
8. transformer thermometer according to claim 1 detection and recognition methods, which is characterized in that the step 6 includes,
Extract the profile point set of the pointer profile；
It finds in the profile meeting point at a distance of maximum two points, straight line is done with this two o'clock；
Continue to find the point pair on pointer profile, forms it into the line segment vertical with the straight line, and find in from the point
The point pair of lie farthest away；
The point of the lie farthest away is the center of circle of the transformer thermometer to the line segment of formation and the point of the straight line intersection,
Coordinate system is established by origin of the center of circle；
Based on any two points in profile meeting point in the vector and coordinate system that maximum two points are formed in vertical axis to
Amount calculates the reading of pointer direction of rotation and direction using the cosine law.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910014109.3A CN109784257B (en) | 2019-01-08 | 2019-01-08 | Transformer thermometer detection and identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910014109.3A CN109784257B (en) | 2019-01-08 | 2019-01-08 | Transformer thermometer detection and identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109784257A true CN109784257A (en) | 2019-05-21 |
| CN109784257B CN109784257B (en) | 2021-10-12 |
Family
ID=66499158
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910014109.3A Active CN109784257B (en) | 2019-01-08 | 2019-01-08 | Transformer thermometer detection and identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109784257B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111311580A (en) * | 2020-02-19 | 2020-06-19 | 中冶赛迪重庆信息技术有限公司 | Steam drum liquid level abnormity identification method and system based on image identification |
| CN113780263B (en) * | 2021-09-03 | 2023-06-16 | 华南师范大学 | Method and device for positioning and identifying reading of pressure alarm instrument |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104751187A (en) * | 2015-04-14 | 2015-07-01 | 山西科达自控股份有限公司 | Automatic meter-reading image recognition method |
| CN105868776A (en) * | 2016-03-25 | 2016-08-17 | 中国科学院自动化研究所 | Transformer equipment recognition method and device based on image processing technology |
| US20180130174A1 (en) * | 2016-11-10 | 2018-05-10 | Samsung Display Co., Ltd. | Display apparatus, controlling method thereof, and terminal thereof |
| CN108491838A (en) * | 2018-03-08 | 2018-09-04 | 南京邮电大学 | Pointer-type gauges registration read method based on SIFT and HOUGH |
| CN108764257A (en) * | 2018-05-23 | 2018-11-06 | 郑州金惠计算机系统工程有限公司 | A kind of pointer instrument recognition methods of various visual angles |
| CN108764134A (en) * | 2018-05-28 | 2018-11-06 | 江苏迪伦智能科技有限公司 | A kind of automatic positioning of polymorphic type instrument and recognition methods suitable for crusing robot |
| CN109447062A (en) * | 2018-09-29 | 2019-03-08 | 南京理工大学 | Pointer-type gauges recognition methods based on crusing robot |
-
2019
- 2019-01-08 CN CN201910014109.3A patent/CN109784257B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104751187A (en) * | 2015-04-14 | 2015-07-01 | 山西科达自控股份有限公司 | Automatic meter-reading image recognition method |
| CN105868776A (en) * | 2016-03-25 | 2016-08-17 | 中国科学院自动化研究所 | Transformer equipment recognition method and device based on image processing technology |
| US20180130174A1 (en) * | 2016-11-10 | 2018-05-10 | Samsung Display Co., Ltd. | Display apparatus, controlling method thereof, and terminal thereof |
| CN108491838A (en) * | 2018-03-08 | 2018-09-04 | 南京邮电大学 | Pointer-type gauges registration read method based on SIFT and HOUGH |
| CN108764257A (en) * | 2018-05-23 | 2018-11-06 | 郑州金惠计算机系统工程有限公司 | A kind of pointer instrument recognition methods of various visual angles |
| CN108764134A (en) * | 2018-05-28 | 2018-11-06 | 江苏迪伦智能科技有限公司 | A kind of automatic positioning of polymorphic type instrument and recognition methods suitable for crusing robot |
| CN109447062A (en) * | 2018-09-29 | 2019-03-08 | 南京理工大学 | Pointer-type gauges recognition methods based on crusing robot |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111311580A (en) * | 2020-02-19 | 2020-06-19 | 中冶赛迪重庆信息技术有限公司 | Steam drum liquid level abnormity identification method and system based on image identification |
| CN113780263B (en) * | 2021-09-03 | 2023-06-16 | 华南师范大学 | Method and device for positioning and identifying reading of pressure alarm instrument |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109784257B (en) | 2021-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112818988B (en) | Automatic identification reading method and system for pointer instrument | |
| CN102521560B (en) | Instrument pointer image identification method of high-robustness rod | |
| CN109583324A (en) | A kind of pointer meters reading automatic identifying method based on the more box detectors of single-point | |
| CN106529559A (en) | Pointer-type circular multi-dashboard real-time reading identification method | |
| CN105023008A (en) | Visual saliency and multiple characteristics-based pedestrian re-recognition method | |
| CN112598733B (en) | Ship detection method based on multi-mode data fusion compensation adaptive optimization | |
| CN109447061A (en) | Reactor oil level indicator recognition methods based on crusing robot | |
| CN109447062A (en) | Pointer-type gauges recognition methods based on crusing robot | |
| CN103034838A (en) | Special vehicle instrument type identification and calibration method based on image characteristics | |
| CN108491838A (en) | Pointer-type gauges registration read method based on SIFT and HOUGH | |
| CN111563896B (en) | Image processing method for detecting abnormality of overhead line system | |
| CN108960115A (en) | Multi-direction Method for text detection based on angle point | |
| CN103854278A (en) | Printed circuit board image registration method based on shape context of mass center of communicated region | |
| CN109389165A (en) | Oil level gauge for transformer recognition methods based on crusing robot | |
| CN112288758B (en) | Infrared and visible light image registration method for power equipment | |
| CN109784257A (en) | A kind of detection of transformer thermometer and recognition methods | |
| CN106709523B (en) | Optical remote sensing image ship identification method based on S-HOG characteristics | |
| CN114821358A (en) | Optical remote sensing image marine ship target extraction and identification method | |
| CN109858474A (en) | A kind of detection of transformer oil surface temperature controller and recognition methods | |
| CN106682668A (en) | Power transmission line geological disaster monitoring method using unmanned aerial vehicle to mark images | |
| US20230386023A1 (en) | Method for detecting medical images, electronic device, and storage medium | |
| CN116310263A (en) | Pointer type aviation horizon instrument indication automatic reading implementation method | |
| CN116385477A (en) | Tower image registration method based on image segmentation | |
| CN107220612B (en) | Fuzzy face discrimination method taking high-frequency analysis of local neighborhood of key points as core | |
| CN109360289B (en) | Power meter detection method fusing inspection robot positioning information |
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 |