CN109214398B - Method and system for measuring rod position from continuous images - Google Patents
Method and system for measuring rod position from continuous images Download PDFInfo
- Publication number
- CN109214398B CN109214398B CN201811096454.8A CN201811096454A CN109214398B CN 109214398 B CN109214398 B CN 109214398B CN 201811096454 A CN201811096454 A CN 201811096454A CN 109214398 B CN109214398 B CN 109214398B
- Authority
- CN
- China
- Prior art keywords
- rod body
- measured
- image
- representing
- rod
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 83
- 238000005259 measurement Methods 0.000 claims abstract description 67
- 238000012545 processing Methods 0.000 claims abstract description 41
- 238000004364 calculation method Methods 0.000 claims abstract description 23
- 238000003672 processing method Methods 0.000 claims abstract description 20
- 210000000746 body region Anatomy 0.000 claims description 20
- 238000007781 pre-processing Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 8
- 230000009467 reduction Effects 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 description 5
- 238000013178 mathematical model Methods 0.000 description 5
- 238000007796 conventional method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method and a system for measuring the position of a rod body from continuous images, wherein the method comprises the following steps: acquiring at least two input images of the rod body to be measured, wherein the position of the rod body to be measured is recorded; classifying images of the input images, identifying and segmenting rod body areas corresponding to the input images, and performing corresponding image processing on the images of the rod body areas according to a preset processing method to obtain initial coordinate points for representing the positions of the rod bodies to be measured; and acquiring each track information of the rod body to be measured and each initial coordinate point for representing the position of the rod body to be measured, and performing corresponding measurement calculation processing on the image of each rod body area according to each track information and each initial coordinate point to obtain a corresponding vertex space coordinate for representing the vertex position of the rod body to be measured. The scheme of the invention has the advantages of low cost for acquiring the data source, small data quantity of the required data source, full automation of the measuring method and high measuring efficiency.
Description
Technical Field
The invention relates to the technical field of measurement, in particular to a method and a system for measuring the position of a rod body from continuous images.
Background
In the existing method for acquiring the position information of the rod body to be positioned, a common method is to directly acquire the position information of the rod body to be positioned from point cloud data.
Before the position information of the to-be-measured rod body is directly acquired from the point cloud data, a data system comprising the point cloud data needs to be established, the point cloud data system comprising the position data of each to-be-measured rod body needs to be established within a preset measuring range, time and labor are consumed, a large number of large advanced devices are needed to technically support the technical support, the cost for purchasing the large number of large advanced devices is high, and for the situation that the position data of each to-be-measured rod body within the preset measuring range only needs to be accurately acquired, the process for establishing the data system of the point cloud data is complex, the cost is too high, universality is not achieved, and the general economic income of users cannot be met.
The existing method for acquiring the position information of the rod to be positioned has the following defects: the method has high cost and no universality and is a technical problem to be solved.
Disclosure of Invention
An embodiment of the present invention provides a method and a system for measuring a position of a stick from consecutive images, so as to solve the problems that the cost for obtaining the position of the stick is high and the obtaining method of the position information of the stick in the prior art is not universal.
To achieve the above object, an embodiment of the present invention provides a method for measuring a position of a rod from consecutive images, the method including: acquiring at least two continuous input images of the rod body to be measured, wherein the position of the rod body to be measured is recorded; classifying images of the input images, identifying and segmenting rod body regions corresponding to the input images, and performing corresponding image processing on the images of the rod body regions according to a preset processing method to obtain initial coordinate points for representing the positions of the rod bodies to be measured; and acquiring each track information of the rod body to be measured and each initial coordinate point for representing the position of the rod body to be measured, and performing corresponding measurement calculation processing on the image of each rod body area according to each track information and each initial coordinate point to obtain a corresponding vertex space coordinate for representing the vertex position of the rod body to be measured.
Optionally, the image classification of each input image, identifying and segmenting each rod body region corresponding to each input image, and performing image processing on the image of each rod body region according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured specifically includes: classifying the input images, and identifying and segmenting rod body areas corresponding to the input images; extracting each characteristic point of each rod body from each rod body area corresponding to each input image, calculating pixels of each characteristic point, performing characteristic point matching according to the pixels of each characteristic point to obtain a corresponding characteristic point matching result, and performing target tracking on the rod body to be measured according to the corresponding characteristic point matching result to obtain a corresponding rod body target continuous frame of the rod body to be measured; selecting a farthest measuring frame and a nearest measuring frame from target continuous frames of each rod body according to a first preset rule, and performing measuring calculation according to matching homonymous points between the farthest measuring frame and the nearest measuring frame to obtain a group of space coordinates capable of representing the position of the rod body to be measured; and performing RANSAC processing on the corresponding group of space coordinates according to the preset processing method to obtain corresponding initial coordinate points for representing the position of the rod body to be measured.
Optionally, the obtaining of each track information of the to-be-measured rod body and each initial coordinate point for representing the position of the to-be-measured rod body, and performing corresponding image processing on the image of each rod body area according to each track information and each initial coordinate point to obtain a corresponding vertex space coordinate for representing the vertex position of the to-be-measured rod body specifically includes: acquiring track information of the rod body to be measured and initial coordinate points for representing the position of the rod body to be measured; obtaining a corresponding projection plane of the rod body to be measured by using a point method according to the track information and the initial coordinate points of the position of the rod body to be measured; searching the pixel coordinates of the vertex of the rod body to be measured in each rod body area at the division position; and obtaining corresponding vertex space coordinates for representing the vertex position of the rod body to be measured according to the corresponding pixel coordinates of the vertex of the rod body to be measured and a line-plane intersection calculation method.
Optionally, after the obtaining of the at least two input images of the rod to be measured, in which the positions of the rod to be measured are recorded, the method further includes: preprocessing each input image, wherein the preprocessing of each input image specifically comprises: and carrying out binarization processing on each input image, or identifying a corresponding preprocessing area for each input image.
Optionally, after identifying and segmenting each rod region corresponding to each input image, the method further includes: and carrying out noise reduction processing on each rod body area corresponding to each input image.
Optionally, the method further comprises: sequencing the sequence of the images continuously existing in the rod body to be measured according to a second preset rule, wherein the second preset rule is that in the images continuously existing in the rod body to be measured, the more the current image sequence is sequenced, the clearer the rod body outline of the rod body to be measured in the corresponding image is.
Optionally, the first preset rule is that, in each image in which the rod to be measured continuously exists, a frame in which the rod profile of the rod to be measured completely appears in the image for the first time is used as a first measurement frame, the first measurement frame is a farthest measurement frame, the first measurement frame is used as a reference frame, a second measurement frame which is a preset distance away from the first measurement frame is calculated, the second measurement frame is a closest measurement frame, the rod profile of the rod to be measured in the second measurement frame is clear and complete, and a distance value corresponding to the preset distance is greater than zero.
Optionally, the preset processing method uses a RANSAC algorithm to process the corresponding group of spatial coordinates, so as to obtain each initial coordinate point for representing the position of the rod body to be measured.
Optionally, verifying the calculated position of the rod body to be measured by a back projection method; and the verifying the calculated position of the rod body to be measured by the back projection method specifically comprises the following steps: verifying each calculated initial coordinate point of the rod body to be measured by a back projection method; and/or verifying the calculated space coordinates of each vertex of the rod body to be measured by a back projection method.
In keeping with the above method, according to another aspect of the invention, there is provided a system for measuring the position of a shaft from successive images, the system comprising: the acquisition unit is used for acquiring at least two continuous input images of the rod body to be measured, wherein the position of the rod body to be measured is recorded; acquiring track information of the rod body to be measured and initial coordinate points for representing the position of the rod body to be measured; the processing unit is used for carrying out image classification on each input image acquired by the acquisition unit, identifying and segmenting each rod body area corresponding to each input image, and carrying out corresponding image processing on the image of each rod body area according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured; and performing corresponding measurement calculation processing on the image of each rod body area according to each track information and each initial coordinate point acquired by the acquisition unit to obtain a corresponding vertex space coordinate for representing the vertex position of the rod body to be measured.
The embodiment of the invention has the following advantages: compared with the conventional method for acquiring the position of the rod body from the point cloud data, the method and the system for measuring the position of the rod body from the continuous image have the advantages of low cost for acquiring a data source, small data quantity of the required data source, full automation of the measuring method and high measuring efficiency.
Drawings
Fig. 1 is a schematic flow chart of a method for measuring a position of a rod from consecutive images according to embodiment 1 of the present invention;
FIG. 2 is a schematic flow chart of a method for measuring a position of a rod from consecutive images according to embodiment 2 of the present invention;
fig. 3 is a schematic structural diagram of a system for measuring a position of a rod from consecutive images according to embodiment 3 of the present invention.
In the embodiments of the invention, the reference numbers are as follows, in combination with the attached drawings:
301-an acquisition unit; 302-processing unit.
Detailed Description
The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
Example 1
According to an embodiment of the present invention, a method for measuring a position of a rod from consecutive images is provided, as shown in fig. 1, which is a flow chart of a method for measuring a position of a rod from consecutive images provided in embodiment 1 of the present invention. The method for measuring the position of the light pole in the continuous image, provided by the embodiment 1 of the invention, is an image vision-based measurement algorithm for the position of the light pole near a road, and comprises the following steps:
s101, acquiring at least two continuous input images of the rod body to be measured, wherein the position of the rod body to be measured is recorded;
s102, classifying images of the input images, identifying and segmenting rod body regions corresponding to the input images, and performing corresponding image processing on the images of the rod body regions according to a preset processing method to obtain initial coordinate points for representing the positions of rod bodies to be measured;
s103, acquiring each track information of the rod body to be measured and each initial coordinate point for representing the position of the rod body to be measured, and performing corresponding measurement calculation processing on the image of each rod body area according to each track information and each initial coordinate point to obtain a corresponding vertex space coordinate for representing the vertex position of the rod body to be measured; therefore, compared with the existing method for acquiring the rod position information from the point cloud data, the method for measuring the rod position provided by embodiment 1 of the invention has the advantages of low cost for acquiring the data source, small data quantity of the required data source, full automation of the measuring method and high measuring efficiency.
In an optional example, the image classification of each input image, the identification and segmentation of each rod body region corresponding to each input image, and the image processing of each rod body region according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured specifically include: classifying the input images, and identifying and segmenting rod body areas corresponding to the input images; extracting each characteristic point of each rod body from each rod body area corresponding to each input image, calculating pixels of each characteristic point, performing characteristic point matching according to the pixels of each characteristic point to obtain a corresponding characteristic point matching result, and performing target tracking on the rod body to be measured according to the corresponding characteristic point matching result to obtain a rod body target continuous frame of the corresponding rod body to be measured; selecting a farthest measuring frame and a nearest measuring frame from target continuous frames of each rod body according to a first preset rule, and performing measuring calculation according to matching homonymous points between the farthest measuring frame and the nearest measuring frame to obtain a group of space coordinates capable of representing the position of the rod body to be measured; performing RANSAC processing on the corresponding group of spatial coordinates according to a preset processing method to obtain corresponding spatial coordinates of each initial coordinate point for representing the position of the rod body to be measured; obtaining initial coordinate points for representing the positions of the rod bodies to be measured through the process of measuring, calculating and processing the images of the rod body areas; therefore, compared with the existing method for acquiring the rod body position information from the point cloud data, the method for measuring the rod body position provided by the embodiment of the invention has the advantages of low cost for acquiring the data source and small data quantity of the required data source, and can measure each corresponding initial coordinate point for representing the rod body position to be measured from the continuous image.
It should be noted that the height value in the space coordinate of each initial coordinate point is not the actual height of the stick body, and in the actual measurement calculation process, the height value in the space coordinate needs to be correspondingly processed to obtain the actual height value of the stick body.
In addition, the method for measuring the position of the rod body is completely automatic, manual operation and recording of measurement data are not needed, and therefore the measurement efficiency is high.
It should be noted that, in the preset processing method in step S102, a RANSAC algorithm is used to process a corresponding set of spatial coordinates, so as to obtain initial coordinate points representing the positions of the rod bodies to be measured.
RANSAC is an abbreviation of Random Sample Consensus, and is an algorithm for calculating mathematical model parameters of data according to a group of Sample data sets containing abnormal data to obtain effective Sample data.
The RANSAC algorithm is often used in computer vision. For example, the matching point problem of a pair of cameras and the calculation of a fundamental matrix are simultaneously solved in the field of stereoscopic vision.
The basic assumption of the RANSAC algorithm is that samples contain correct data (inliers, data that can be described by a model) and also contain abnormal data (outliers, data that is far from a normal range and cannot adapt to a mathematical model), that is, data sets contain noise. These outlier data may be due to erroneous measurements, erroneous assumptions, erroneous calculations, etc. RANSAC also assumes that, given a correct set of data, there is a way to calculate the model parameters that fit into the data.
The RANSAC-based algorithm is a conventional technique and will not be described herein.
In practical application, the sequence of the images of the rod body to be measured continuously exists is sorted according to a second preset rule, wherein the second preset rule is that in the images of the rod body to be measured continuously, the more the current image sequence is sorted, the clearer the rod body outline of the rod body to be measured in the corresponding image is.
It should be noted that, in each image in which the rods to be measured continuously exist, the first predetermined rule is that, taking a frame in which the rod profile of the rod to be measured completely appears in the image for the first time as a first measurement frame, the first measurement frame being the farthest measurement frame, and taking the first measurement frame as a reference frame, calculating a second measurement frame which is a predetermined distance away from the first measurement frame, the second measurement frame being the closest measurement frame, and the rod profile of the rod to be measured in the second measurement frame being clear and complete, where a distance value corresponding to the predetermined distance is greater than zero.
It should be noted that, in practical applications, the first measurement frame (the farthest measurement frame) and the second measurement frame (the nearest measurement frame) cannot be the same measurement frame.
In an optional example, the obtaining of each piece of trajectory information of the to-be-measured rod body and each initial coordinate point for representing the position of the to-be-measured rod body, and performing corresponding image processing on the image of each rod body region according to each piece of trajectory information and each initial coordinate point to obtain a corresponding vertex space coordinate for representing the vertex position of the to-be-measured rod body specifically includes: acquiring track information of a rod body to be measured and initial coordinate points for representing the position of the rod body to be measured; obtaining a projection plane of the corresponding rod body to be measured by using a point method according to the track information and the initial coordinate points of the position of the rod body to be measured; searching the pixel coordinates of the vertex of the corresponding rod body to be measured in each rod body area at the division position; obtaining corresponding vertex space coordinates for representing the vertex position of the rod body to be measured according to the corresponding pixel coordinates of the vertex of the rod body to be measured and a line-plane intersection calculation method; therefore, compared with the existing method for acquiring the rod body position information from the point cloud data, the method for measuring the rod body position provided by the embodiment of the invention has the advantages of low cost for acquiring the data source and small data quantity of the required data source, and can measure the corresponding vertex space coordinate for representing the vertex position of the rod body to be measured from the continuous image.
In addition, the method for measuring the vertex position of the rod body is completely automatic, manual operation and recording of measurement data are not needed, and therefore the measurement efficiency is high.
In an optional example, after acquiring at least two input images of the stick to be measured recording the position of the stick to be measured, the method further comprises: the preprocessing of each input image, and the preprocessing of each input image specifically includes: and carrying out binarization processing on each input image, or identifying a corresponding preprocessing area for each input image.
Optionally, after identifying and segmenting each rod region corresponding to each input image, the method further includes: carrying out noise reduction processing on each rod body region corresponding to each input image; therefore, interference is avoided by carrying out noise reduction processing on each input image, and each processed input image is clearer.
In an optional example, the calculated position of the rod body to be measured is verified through a back projection method; and verifying the calculated position of the rod body to be measured by a back projection method specifically comprises the following steps: verifying each calculated initial coordinate point of the rod body to be measured by a back projection method; and/or verifying the calculated spatial coordinates of each vertex of the rod body to be measured by a back projection method; thus, each initial coordinate point of the rod body to be measured is verified through the back projection method, and each vertex space coordinate of the rod body to be measured is verified, so that whether each initial coordinate point and each vertex space coordinate which are related to the position of the rod body to be measured and calculated are correct or not is verified, whether the initial coordinate points and the vertex space coordinates are in a corresponding preset error range or not is verified, and if the verification result shows that: and outputting and displaying to a user that each initial coordinate point and each vertex space coordinate associated with the position of the rod body to be measured are within a corresponding preset error range: and (4) calculating the initial coordinate points and the vertex space coordinates of the position of the rod body to be measured, otherwise, recalculating the position of the rod body to be measured, and repeating the calculation process until the position of the rod body to be measured is finally calculated. It should be noted that the back-projection method is a conventional technical means, and is not described in detail in embodiment 1 of the present invention.
Example 2
According to an embodiment of the present invention, a method for measuring a position of a rod from a continuous image is provided, and as shown in fig. 2, another flow chart of the method for measuring a position of a rod from a continuous image is provided for embodiment 2 of the present invention. The method for measuring the position of the light pole in the continuous image, which is provided by the embodiment 2 of the invention, is an image vision-based measuring algorithm for the position of the light pole near the road, and the method comprises the following steps:
step 1, preprocessing an input image, including establishing an interested region, binaryzation and the like.
And 2, classifying the images, identifying the rod, and then segmenting the rod body area.
And 3, carrying out noise reduction treatment (corrosion collision and the like) on the rod body area.
And 4, extracting rod body characteristic points in the image, calculating pixels and descriptors of the point characteristic points in the image, matching the characteristic points, and tracking a target (in which continuous frames a certain rod body appears) according to a matching result.
And 5, selecting an optimal measuring frame, taking a frame which completely appears in the image for the first time as a nearest measuring frame in the continuous rod body images (the rod body outline in the images which are sequenced in sequence and are in front of the sequence is clearer and the measuring effect is better), and calculating a frame at a certain distance ahead by taking the closest measuring frame as a reference, so as to find out a farthest measuring frame, and simultaneously judging whether the rod body outline in the frame is complete, otherwise, judging the last frame until the rod body outline in the frame is complete and not the nearest measuring frame.
And 6, carrying out forward intersection measurement calculation on the matched characteristic point pairs in the nearest measurement frame and the farthest measurement frame to obtain a group of space coordinates, wherein the group of space coordinates represent the position on the rod body.
And 7, conducting RANSAC processing on the set of rod body position coordinates calculated in the step 6 to obtain a coordinate point representing the position of the rod body.
It should be noted that RANSAC is an abbreviation of Random Sample Consensus, and is an algorithm for calculating mathematical model parameters of data according to a set of Sample data sets containing abnormal data to obtain valid Sample data.
The RANSAC algorithm is often used in computer vision. For example, the matching point problem of a pair of cameras and the calculation of a fundamental matrix are simultaneously solved in the field of stereoscopic vision.
The basic assumption of the RANSAC algorithm is that samples contain correct data (inliers, data that can be described by a model) and also contain abnormal data (outliers, data that is far from a normal range and cannot adapt to a mathematical model), that is, data sets contain noise. These outlier data may be due to erroneous measurements, erroneous assumptions, erroneous calculations, etc. RANSAC also assumes that, given a correct set of data, there is a way to calculate the model parameters that fit into the data.
The RANSAC-based algorithm is a conventional technique and will not be described herein.
And 8, obtaining a projection plane of the rod body by using a point method through the acquisition track of the image and the coordinates of the rod body.
And 9, finding the pixel coordinates of the rod body vertex in the rod body area at the division position.
And 10, obtaining the space coordinate position of the rod body vertex through a calculation mode of line-surface intersection.
And 11, verifying the calculated rod body position through back projection.
Compared with the existing method for acquiring the position information of the rod body, the method for measuring the position of the rod body from the continuous images provided by the embodiment 2 of the invention has the following advantages:
1. the method for measuring the position of the rod body provided by the embodiment 2 of the invention has low cost for acquiring the data source and small data quantity of the required data source.
2. The method for measuring the position of the rod body provided by the embodiment 2 of the invention is fully automatic, and the measuring efficiency of the measuring method for measuring the position of the rod body is high.
Example 3
According to an embodiment of the present invention, there is also provided a system for measuring a position of a rod from consecutive images, as shown in fig. 3, which is a schematic structural diagram of a system for measuring a position of a rod from consecutive images according to embodiment 3 of the present invention.
The system for measuring the position of the rod body from the continuous images provided by embodiment 3 of the present invention includes an acquisition unit 301 and a processing unit 302.
Specifically, the obtaining unit 301 obtains at least two continuous input images of the rod to be measured, in which the position of the rod to be measured is recorded; and
acquiring track information of a rod body to be measured and initial coordinate points for representing the position of the rod body to be measured;
the processing unit 302 is configured to perform image classification on each input image acquired by the acquisition unit 301, identify and segment each rod body region corresponding to each input image, and perform corresponding image processing on the image of each rod body region according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured; and
performing corresponding measurement calculation processing on the image of each rod body area according to each track information and each initial coordinate point acquired by the acquisition unit 301 to obtain a corresponding vertex space coordinate for representing the vertex position of the rod body to be measured; thus, compared with the conventional system for acquiring the position information of the rod from the point cloud data, the system for measuring the position of the rod provided by embodiment 3 of the present invention has the advantages of low cost for acquiring the data source, small data amount of the required data source, fully automatic measuring method, and high measuring efficiency.
In an optional example, the processing unit 302 is specifically configured to: classifying the input images, and identifying and segmenting rod body areas corresponding to the input images; extracting each characteristic point of each rod body from each rod body area corresponding to each input image, calculating pixels of each characteristic point, performing characteristic point matching according to the pixels of each characteristic point to obtain a corresponding characteristic point matching result, and performing target tracking on the rod body to be measured according to the corresponding characteristic point matching result to obtain a rod body target continuous frame of the corresponding rod body to be measured; selecting a farthest measuring frame and a nearest measuring frame from target continuous frames of each rod body according to a first preset rule, and performing measuring calculation according to matching homonymous points between the farthest measuring frame and the nearest measuring frame to obtain a group of space coordinates capable of representing the position of the rod body to be measured; performing RANSAC processing on the corresponding group of space coordinates according to a preset processing method to obtain corresponding initial coordinate points for representing the position of the rod body to be measured; thus, compared with the existing method for acquiring the rod position information from the point cloud data, the system for measuring the rod position provided by embodiment 3 of the present invention has the advantages of low cost for acquiring the data source and small data amount of the required data source, and can measure each initial coordinate point for representing the rod position to be measured from the continuous image.
In addition, the system for measuring the position of the rod body is completely automatic, manual operation and recording of measurement data are not needed, and therefore the measurement efficiency is high.
It should be noted that, in the preset processing method in the system for measuring the position of the rod according to embodiment 3 of the present invention, a RANSAC algorithm is used to process a corresponding set of spatial coordinates, so as to obtain each initial coordinate point for representing the position of the rod to be measured.
RANSAC is an abbreviation of Random Sample Consensus, and is an algorithm for calculating mathematical model parameters of data according to a group of Sample data sets containing abnormal data to obtain effective Sample data. For the rest of the relevant description about RANSAC, please refer to the relevant description in embodiment 1, and the RANSAC-based algorithm is a conventional technique and will not be described herein.
In practical application, the sequence of the images of the rod body to be measured continuously exists is sorted according to a second preset rule, wherein the second preset rule is that in the images of the rod body to be measured continuously, the more the current image sequence is sorted, the clearer the rod body outline of the rod body to be measured in the corresponding image is.
It should be noted that, in each image in which the rods to be measured continuously exist, the first predetermined rule is that, taking a frame in which the rod profile of the rod to be measured completely appears in the image for the first time as a first measurement frame, the first measurement frame being the farthest measurement frame, and taking the first measurement frame as a reference frame, calculating a second measurement frame which is a predetermined distance away from the first measurement frame, the second measurement frame being the closest measurement frame, and the rod profile of the rod to be measured in the second measurement frame being clear and complete, where a distance value corresponding to the predetermined distance is greater than zero.
It should be noted that, in practical applications, the first measurement frame (the farthest measurement frame) and the second measurement frame (the nearest measurement frame) cannot be the same measurement frame.
Although the invention has been described in detail above with reference to a general description and specific examples, it will be apparent to one skilled in the art that modifications or improvements may be made thereto based on the invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.
Claims (9)
1. A method of measuring the position of a shaft from successive images, comprising:
acquiring at least two continuous input images of the rod body to be measured, wherein the position of the rod body to be measured is recorded;
classifying images of the input images, identifying and segmenting rod body regions corresponding to the input images, and performing corresponding image processing on the images of the rod body regions according to a preset processing method to obtain initial coordinate points for representing the positions of the rod bodies to be measured;
acquiring track information of the rod body to be measured and initial coordinate points used for representing the position of the rod body to be measured, and performing corresponding measurement calculation processing on the image of each rod body area according to the track information and the initial coordinate points to obtain corresponding vertex space coordinates used for representing the vertex position of the rod body to be measured;
the image classification of each input image, the identification and segmentation of each rod body region corresponding to each input image, and the image processing of each rod body region according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured specifically include:
classifying the input images, and identifying and segmenting rod body areas corresponding to the input images;
extracting each characteristic point of each rod body from each rod body area corresponding to each input image, calculating pixels of each characteristic point, performing characteristic point matching according to the pixels of each characteristic point to obtain a corresponding characteristic point matching result, and performing target tracking on the rod body to be measured according to the corresponding characteristic point matching result to obtain a corresponding rod body target continuous frame of the rod body to be measured;
selecting a farthest measuring frame and a nearest measuring frame from target continuous frames of each rod body according to a first preset rule, and performing measuring calculation according to matching homonymous points between the farthest measuring frame and the nearest measuring frame to obtain a group of space coordinates capable of representing the position of the rod body to be measured;
and performing RANSAC processing on the corresponding group of space coordinates according to the preset processing method to obtain corresponding initial coordinate points for representing the position of the rod body to be measured.
2. The method according to claim 1, wherein the obtaining of each piece of trajectory information of the to-be-measured rod body and each initial coordinate point for representing the position of the to-be-measured rod body, and performing corresponding image processing on the image of each rod body region according to each piece of trajectory information and each initial coordinate point to obtain a corresponding vertex space coordinate for representing the vertex position of the to-be-measured rod body specifically comprises:
acquiring track information of the rod body to be measured and initial coordinate points for representing the position of the rod body to be measured;
obtaining a corresponding projection plane of the rod body to be measured by using a point method according to the track information and the initial coordinate points of the position of the rod body to be measured;
searching the pixel coordinates of the vertex of the rod body to be measured in each rod body area at the division position;
and obtaining corresponding vertex space coordinates for representing the vertex position of the rod body to be measured according to the corresponding pixel coordinates of the vertex of the rod body to be measured and a line-plane intersection calculation method.
3. The method of claim 1, wherein after said obtaining at least two input images of a stick to be measured recording positions of the stick to be measured, the method further comprises:
preprocessing each input image, and
the preprocessing of each input image specifically comprises:
each input image is subjected to binarization processing, or,
a corresponding pre-processing region is identified for each input image.
4. The method of claim 1, wherein after identifying and segmenting the respective stick body regions corresponding to the respective input images, the method further comprises:
and carrying out noise reduction processing on each rod body area corresponding to each input image.
5. The method of claim 1, further comprising:
sequencing the sequence of the images continuously existing in the rod body to be measured according to a second preset rule, wherein the second preset rule is that in the images continuously existing in the rod body to be measured, the more the current image sequence is sequenced, the clearer the rod body outline of the rod body to be measured in the corresponding image is.
6. The method of claim 5,
the first preset rule is that, in each image in which the rod to be measured continuously exists, a frame in which the rod profile of the rod to be measured completely appears in the image for the first time is used as a first measurement frame, the first measurement frame is a farthest measurement frame, the first measurement frame is used as a reference frame, a second measurement frame which is a preset distance away from the first measurement frame is calculated, the second measurement frame is a closest measurement frame, the rod profile of the rod to be measured in the second measurement frame is clear and complete, and a distance value corresponding to the preset distance is greater than zero.
7. The method of claim 1,
the preset processing method adopts RANSAC algorithm to process a group of corresponding space coordinates to obtain each initial coordinate point for representing the position of the rod body to be measured.
8. The method of claim 1,
verifying the calculated position of the rod body to be measured by a back projection method; and
the verifying the calculated position of the rod body to be measured by the back projection method specifically comprises the following steps:
verifying each calculated initial coordinate point of the rod body to be measured by a back projection method; and/or verifying the calculated space coordinates of each vertex of the rod body to be measured by a back projection method.
9. A system for measuring the position of a shaft from successive images, comprising:
the acquisition unit is used for acquiring at least two continuous input images of the rod body to be measured, wherein the position of the rod body to be measured is recorded; and
acquiring track information of the rod body to be measured and initial coordinate points for representing the position of the rod body to be measured;
the processing unit is used for carrying out image classification on each input image acquired by the acquisition unit, identifying and segmenting each rod body area corresponding to each input image, and carrying out corresponding image processing on the image of each rod body area according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured; and
performing corresponding measurement calculation processing on the image of each rod body area according to each track information and each initial coordinate point acquired by the acquisition unit to obtain a corresponding vertex space coordinate for representing the vertex position of the rod body to be measured;
the image classification of each input image, the identification and segmentation of each rod body region corresponding to each input image, and the image processing of each rod body region according to a preset processing method to obtain each initial coordinate point for representing the position of the rod body to be measured specifically include:
classifying the input images, and identifying and segmenting rod body areas corresponding to the input images;
extracting each characteristic point of each rod body from each rod body area corresponding to each input image, calculating pixels of each characteristic point, performing characteristic point matching according to the pixels of each characteristic point to obtain a corresponding characteristic point matching result, and performing target tracking on the rod body to be measured according to the corresponding characteristic point matching result to obtain a corresponding rod body target continuous frame of the rod body to be measured;
selecting a farthest measuring frame and a nearest measuring frame from target continuous frames of each rod body according to a first preset rule, and performing measuring calculation according to matching homonymous points between the farthest measuring frame and the nearest measuring frame to obtain a group of space coordinates capable of representing the position of the rod body to be measured; and performing RANSAC processing on the corresponding group of space coordinates according to the preset processing method to obtain corresponding initial coordinate points for representing the position of the rod body to be measured.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811096454.8A CN109214398B (en) | 2018-09-19 | 2018-09-19 | Method and system for measuring rod position from continuous images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811096454.8A CN109214398B (en) | 2018-09-19 | 2018-09-19 | Method and system for measuring rod position from continuous images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109214398A CN109214398A (en) | 2019-01-15 |
| CN109214398B true CN109214398B (en) | 2020-09-22 |
Family
ID=64984943
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811096454.8A Active CN109214398B (en) | 2018-09-19 | 2018-09-19 | Method and system for measuring rod position from continuous images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109214398B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110853431A (en) * | 2019-10-23 | 2020-02-28 | 重庆大学 | Human-computer interaction system for drawing force diagram in bar system structure |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5060358B2 (en) * | 2008-03-25 | 2012-10-31 | 株式会社トプコン | Surveying system |
| CN107631782B (en) * | 2017-07-18 | 2020-02-21 | 南京邮电大学 | Water level detection method based on Harris angular point detection |
| CN107506406B (en) * | 2017-08-07 | 2020-10-20 | 中国南方电网有限责任公司超高压输电公司广州局 | Picture archiving method and system for power transmission line tower |
-
2018
- 2018-09-19 CN CN201811096454.8A patent/CN109214398B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109214398A (en) | 2019-01-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110705405B (en) | Target labeling method and device | |
| CN108229475B (en) | Vehicle tracking method, system, computer device and readable storage medium | |
| US11145080B2 (en) | Method and apparatus for three-dimensional object pose estimation, device and storage medium | |
| CN111598913B (en) | Image segmentation method and system based on robot vision | |
| US8699786B2 (en) | 3D model generating apparatus, method and CRM by line pattern imaging | |
| CN110570435B (en) | Method and device for carrying out damage segmentation on vehicle damage image | |
| CN111340749B (en) | Image quality detection method, device, equipment and storage medium | |
| CN108805180B (en) | Target object detection method and device | |
| CN109934873B (en) | Method, device and equipment for acquiring marked image | |
| CN112991374B (en) | Canny algorithm-based edge enhancement method, canny algorithm-based edge enhancement device, canny algorithm-based edge enhancement equipment and storage medium | |
| CN113706472B (en) | Highway pavement disease detection method, device, equipment and storage medium | |
| CN111126393A (en) | Vehicle appearance refitting judgment method and device, computer equipment and storage medium | |
| CN109255801B (en) | Method, device and equipment for tracking edges of three-dimensional object in video and storage medium | |
| CN112991327B (en) | Steel grid welding system, method and terminal equipment based on machine vision | |
| CN109214398B (en) | Method and system for measuring rod position from continuous images | |
| CN107392948B (en) | Image registration method of amplitude-division real-time polarization imaging system | |
| CN117315210A (en) | Image blurring method based on stereoscopic imaging and related device | |
| CN115083008A (en) | Moving object detection method, device, equipment and storage medium | |
| CN111985266A (en) | Scale map determination method, device, equipment and storage medium | |
| CN115471537A (en) | Monocular camera-based moving target distance and height measuring method | |
| CN113255405B (en) | Parking space line identification method and system, parking space line identification equipment and storage medium | |
| CN109766012B (en) | Sight line calculation method and device | |
| CN113450335A (en) | Road edge detection method, road edge detection device and road surface construction vehicle | |
| CN110599456A (en) | Method for extracting specific region of medical image | |
| CN109215068B (en) | Image magnification measuring method and device |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220421 Address after: 550000 No. 818, Kaifa Avenue, Guiyang Economic and Technological Development Zone, Guizhou Province (plant of science and engineering company) Patentee after: GUIZHOU KUANDENG ZHIYUN TECHNOLOGY Co.,Ltd. Address before: 811, 8 / F, 101, 3-8 / F, building 17, rongchuang Road, Chaoyang District, Beijing 100012 Patentee before: KUANDENG (BEIJING) TECHNOLOGY Co.,Ltd. |