CN110610522B - Calibration method and device of image acquisition equipment, readable storage medium and system - Google Patents
Calibration method and device of image acquisition equipment, readable storage medium and system Download PDFInfo
- Publication number
- CN110610522B CN110610522B CN201810615799.3A CN201810615799A CN110610522B CN 110610522 B CN110610522 B CN 110610522B CN 201810615799 A CN201810615799 A CN 201810615799A CN 110610522 B CN110610522 B CN 110610522B
- Authority
- CN
- China
- Prior art keywords
- distortion
- image
- image acquisition
- classification data
- classification
- 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 50
- 238000003860 storage Methods 0.000 title claims abstract description 14
- 238000013527 convolutional neural network Methods 0.000 claims description 52
- 238000012549 training Methods 0.000 claims description 44
- 238000012360 testing method Methods 0.000 claims description 40
- 238000012937 correction Methods 0.000 claims description 25
- 238000012545 processing Methods 0.000 claims description 15
- 241000226585 Antennaria plantaginifolia Species 0.000 claims description 14
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 6
- 230000008569 process Effects 0.000 description 9
- 238000013507 mapping Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000003287 optical effect Effects 0.000 description 5
- 230000002776 aggregation Effects 0.000 description 3
- 238000004220 aggregation Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention provides a calibration method, a calibration device, a computer readable storage medium and a calibration system of image acquisition equipment, wherein the method comprises the following steps: acquiring a distorted image of image acquisition equipment; determining classification data of a distorted image through a pre-trained distortion classifier; and burning the classification data into image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data. The classification data corresponding to the distorted image is determined through the distortion classifier, only the classification data is burnt into the image acquisition equipment, the classification data is the serial number corresponding to the distortion class to which the distorted image belongs, the data volume is small, the time spent on burning the classification data is short, the calibration time is shortened, and the calibration efficiency is improved. And the pre-trained distortion classifier can simultaneously carry out distortion classification on multi-frame distortion images, so that parallel calibration on a plurality of image acquisition devices is realized, batch calibration of the image acquisition devices is realized, and the calibration efficiency of the image acquisition devices is greatly improved.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a calibration method, a calibration device, a readable storage medium and a calibration system of image acquisition equipment.
Background
In the process of producing the image acquisition equipment, due to factors such as environmental errors or assembly errors, the optical path diagram of the image acquisition equipment is distorted, so that the image acquisition equipment needs to be subjected to distortion calibration before the image acquisition equipment is used, and the image acquired by the image acquisition equipment is subjected to distortion correction according to a calibration result.
Currently, in the related art, there is a method for calibrating an image capturing device, including: and acquiring a distortion image on image acquisition equipment, and acquiring distortion correction data by using a proportional mapping algorithm by using a preset square grid image and the acquired distortion image. And burning the acquired distortion correction data into the image acquisition equipment to finish the distortion calibration of the image acquisition equipment.
The distortion correction data obtained by the proportional mapping algorithm is an array with a large data size, and it takes a long time to write the distortion correction data into the image acquisition device, so that the calibration efficiency is low.
Disclosure of Invention
In view of the above, an object of the embodiments of the present invention is to provide a calibration method and apparatus for an image capturing device, so as to solve the above-mentioned problems of calibration by a proportional mapping algorithm, that is, the obtained distortion correction data is an array with a large data size, it takes a long time to write the distortion correction data into the image capturing device, and the calibration efficiency is very low.
In a first aspect, an embodiment of the present invention provides a calibration method for an image capture device, where the method includes:
acquiring a distorted image of image acquisition equipment;
determining classification data of the distorted image through a pre-trained distortion classifier;
burning the classification data into the image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data.
With reference to the first aspect, an embodiment of the present invention provides a first possible implementation manner of the first aspect, where the determining, by a pre-trained distortion classifier, classification data of the distorted image includes:
identifying image features of the distorted image through a pre-trained distortion classifier, wherein the image features comprise an image size, a left internal angle degree of inclination and a pincushion distortion radian;
determining a distortion category to which the distorted image belongs through the distortion classifier according to the image features;
and determining the classification data corresponding to the distortion category as the classification data corresponding to the distorted image.
With reference to the first aspect, an embodiment of the present invention provides a second possible implementation manner of the first aspect, where before the obtaining the classification data of the distorted image by using the pre-trained distortion classifier, the method further includes:
acquiring a training sample image set and a test sample image set;
carrying out classification training on the convolutional neural network according to the sample images included in the training sample image set;
when the training of the convolutional neural network meets a preset condition, carrying out distortion classification on the test images included in the test sample image set through the trained convolutional neural network;
and obtaining a classification error corresponding to the trained convolutional neural network, and determining the trained convolutional neural network as a distortion classifier when the classification error is smaller than a preset threshold value.
With reference to the first possible implementation manner of the first aspect, an embodiment of the present invention provides a third possible implementation manner of the first aspect, where the performing, according to the classification data, a distortion correction process on the distorted image includes:
reading classification data from the image acquisition device;
distortion data and standard distortion parameters corresponding to the classification data are obtained;
calculating distortion parameters corresponding to the image acquisition equipment through an image affine algorithm according to the distortion data and the standard distortion parameters;
and carrying out distortion correction processing on the distorted image according to the distortion parameters.
With reference to the first aspect or any one of the second and third possible implementation manners of the first aspect, an embodiment of the present invention provides a fourth possible implementation manner of the first aspect, where before performing distortion correction processing on the distorted image according to the classification data, the method further includes:
the method comprises the steps of obtaining a standard distortion image corresponding to a distortion category, dividing grids into the standard distortion image, obtaining intersection point coordinates of the grids divided on the standard distortion image, and determining the intersection point coordinates as standard distortion parameters corresponding to the distortion category.
In a second aspect, an embodiment of the present invention provides a calibration apparatus for an image capturing device, where the apparatus includes:
the acquisition module is used for acquiring a distorted image of the image acquisition equipment;
the determining module is used for determining classification data of the distorted image through a pre-trained distortion classifier;
and the burning module is used for burning the classification data into the image acquisition equipment and automatically correcting distortion of the distorted image according to the classification data.
With reference to the second aspect, an embodiment of the present invention provides a first possible implementation manner of the second aspect, where the determining module includes:
the identifying unit is used for identifying the image characteristics of the distorted image through a pre-trained distortion classifier, wherein the image characteristics comprise the image size, the left inner angle degree of inclination and the pincushion distortion radian;
the determining unit is used for determining a distortion category to which the distorted image belongs through the distortion classifier according to the image characteristics; and determining the classification data corresponding to the distortion category as the classification data corresponding to the distorted image.
With reference to the second aspect, an embodiment of the present invention provides a second possible implementation manner of the second aspect, where the apparatus further includes:
the classifier training module is used for acquiring a training sample image set and a test sample image set; carrying out classification training on the convolutional neural network according to the sample images included in the training sample image set; when the training of the convolutional neural network meets a preset condition, carrying out distortion classification on the test images included in the test sample image set through the trained convolutional neural network; and obtaining a classification error corresponding to the trained convolutional neural network, and determining the trained convolutional neural network as a distortion classifier when the classification error is smaller than a preset threshold value.
In a third aspect, an embodiment of the present invention provides a computer-readable storage medium for calibrating an image capturing device, including a processor and a memory for storing processor-executable instructions, where the instructions, when executed by the processor, implement steps including:
acquiring a distorted image of image acquisition equipment;
determining classification data of the distorted image through a pre-trained distortion classifier;
burning the classification data into the image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data.
In a fourth aspect, an embodiment of the present invention provides a system for calibrating an image capturing device, including at least one processor and a memory storing computer-executable instructions, where the processor executes the instructions to implement the steps of the method for calibrating an image capturing device according to the first aspect.
In the method, the device, the readable storage medium and the system provided by the embodiment of the invention, the classification data of the distorted image acquired by the image acquisition equipment is determined through a pre-trained distortion classifier; and burning the classification data into image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data. The classification data corresponding to the distorted image is determined through the distortion classifier, only the classification data is burnt into the image acquisition equipment, the classification data is the serial number corresponding to the distortion class to which the distorted image belongs, the data volume is small, the time spent on burning the classification data is short, the calibration time is shortened, and the calibration efficiency is improved. And the pre-trained distortion classifier can simultaneously carry out distortion classification on multi-frame distortion images and can simultaneously carry out parallel calibration on a plurality of image acquisition devices, thereby realizing batch calibration of the image acquisition devices and greatly improving the calibration efficiency of the image acquisition devices.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
FIG. 1 is a flow chart of training a distortion classifier provided in embodiment 1 of the present invention;
fig. 2 is a flowchart illustrating a calibration method for an image capturing device according to embodiment 1 of the present invention;
FIG. 3 is a diagram showing a comparison between before and after image orthotics provided in embodiment 1 of the present invention;
fig. 4 is a schematic diagram illustrating an image distortion correction flow provided in embodiment 1 of the present invention;
fig. 5 shows a schematic structural diagram of a calibration apparatus of an image capturing device according to embodiment 2 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
Considering that the distortion data obtained by the proportional mapping algorithm in the related art is an array with a large data size, it takes a long time to write the distortion data into the image capturing apparatus, and the calibration efficiency is low. Based on this, the embodiment of the present invention provides a calibration method and apparatus for an image capture device, which are described below by way of embodiments.
Example 1
The embodiment of the invention provides a calibration method of image acquisition equipment.
The image capturing device may be an optical sensor such as an optical prism or other device capable of capturing an image. The execution main body of the embodiment of the invention can be an upper computer of the image acquisition equipment. In the production process of the image acquisition equipment, the optical path diagram of the produced image acquisition equipment is usually distorted to a certain degree under the influence of factors such as environmental errors or assembly errors. Before the image acquisition equipment is used, the image acquisition equipment is subjected to distortion calibration by the method provided by the embodiment of the invention, and the calibrated classification data is burnt into the image acquisition equipment, so that the image acquired by the image acquisition equipment is subjected to automatic distortion correction processing according to the classification data.
The classification data is the serial number of the distortion class corresponding to the image acquisition equipment, the data volume of the classification data is very small, and the time spent on burning the classification data into the image acquisition equipment is very short. In addition, the distortion classifier obtained by training in the embodiment of the invention can simultaneously carry out distortion classification on a plurality of image acquisition devices and calibrate the plurality of image acquisition devices, thereby realizing batch calibration of the image acquisition devices and greatly improving the calibration efficiency.
As shown in fig. 1, before calibrating the image capturing device by the method provided by the embodiment of the present invention, the method first trains a distortion classifier through the following operations of steps S1-S5, which specifically includes:
step S1: a training sample image set and a test sample image set are obtained.
In the embodiment of the invention, a plurality of distortion categories are preset, and each distortion category corresponds to different degrees of distortion. For each distortion category, a preset number of sample images are respectively selected, and all the selected sample images form a training sample image set. The preset number may be 10 or 20, etc. The number of sample images included in the training sample image set may be 200 frames or 400 frames, etc.
The more the number of the set distortion categories is, the larger the value of the preset number is, the more the number of the sample images included in the training sample image set is, and the higher the classification accuracy of the distortion classifier trained according to the training sample image set is. The number of the distortion categories and the specific value of the preset number are not specifically limited in the embodiment of the invention, and the number of the distortion categories and the specific value of the preset number can be set according to requirements in practical application.
When a test sample image set is obtained, some image acquisition devices are randomly selected, original images are sampled through the image acquisition devices, and the test images obtained through sampling form the test sample image set.
Step S2: and carrying out classification training on the convolutional neural network according to the sample images included in the training sample image set.
The embodiment of the invention defines the image characteristics of the convolutional neural network needing classification training, wherein the image characteristics comprise the image size, the left internal angle degree of inclination and the pincushion distortion radian. When each distortion category is set, the distortion data corresponding to the distortion category is also set, and the distortion data are the image size, the left inner angle of inclination and the pincushion distortion radian of the distortion image according to the distortion category.
The image sizes range from 445 pixels to 465 pixels on a central vertical axis and 1040 pixels to 1085 pixels on a central horizontal axis. The degree of the inclined left inner angle is in a range of 67 degrees to 71 degrees of the left inner angle of the distorted trapezoid.
A convolutional neural network was constructed by MATLAB (Matrix Laboratory) programming. When the convolutional neural network is subjected to classification training, a sample image is obtained from a training sample image set, grids are divided on the sample image, and the image size, the left internal angle of inclination and the pincushion distortion radian of the sample image are identified through the convolutional neural network. And determining the distortion category of the sample image according to the identified image size, the left inner angle degree of inclination and the pincushion distortion radian.
For each frame of sample image in the training sample image set, the image characteristics of each frame of sample image in the training sample image set are identified through the convolutional neural network according to the mode, the distortion category of each frame of sample image is determined, and the image characteristics corresponding to each distortion category can be learned through the training convolutional neural network based on the classification.
Step S3: and when the training of the convolutional neural network meets the preset condition, carrying out distortion classification on the test images included in the test sample image set through the trained convolutional neural network.
The preset condition may be that a time difference between the current time and the test time of the last test of the convolutional neural network reaches a preset time length, that is, the convolutional neural network trained in step S2 is tested every preset time length.
The preset condition may be that the number of the sample images trained by the convolutional neural network after the last test reaches a preset frame number, that is, each sample image trained by the preset frame number is used for testing the convolutional neural network trained in step S2.
And when the training of the convolutional neural network meets the preset condition, acquiring a test image from the test sample image set, and performing distortion classification on the acquired test image through the convolutional neural network trained in the step S2 at the moment to determine the distortion category to which the test image belongs.
For each frame of test images included in the test sample image set, the distortion category to which each frame of test image belongs can be determined through the convolutional neural network trained in step S2 in the manner described above.
Step S4: and acquiring a classification error corresponding to the trained convolutional neural network, judging whether the classification error is smaller than a preset threshold value, if so, executing the step S5, otherwise, returning to continue executing the step S2.
In the embodiment of the invention, when the test sample image set is obtained, distortion analysis is further performed on each frame of test image included in the test sample image set, and the actual distortion category of each frame of test image is determined. When the classification error corresponding to the trained convolutional neural network is obtained, counting the number of the test images with the distortion types determined by the convolutional neural network and inconsistent with the actual distortion types, calculating the ratio of the counted number to the total frame number of the test images included in the test sample image set, and determining the ratio as the classification error corresponding to the trained convolutional neural network. The classification error is smaller than a preset threshold value, so that the trained convolutional neural network is not excessively converged, the generalization capability is good, and the coverage of the training sample image set and the test sample image set is good.
Step S5: the trained convolutional neural network is determined as a distortion classifier.
In the embodiment of the present invention, the resulting distortion classifier may be a four-layer convolutional neural network, which includes two hidden layers. The size of a sample image included in a training sample image set can be 640 pixels by 480 pixels, the sample image can be expressed as a vector of 640 by 480 pixels, and the data volume of the sample image is large, so that when a convolutional neural network is trained, a local sensing field is set based on spatial relation among parts of the sample image, the local sensing field is used for sensing a local pixel region of the sample image, and a plurality of local pixel regions sensed by the local sensing field are integrated at a higher processing level to obtain global information of the sample image. When the sample image size is 640 × 480 pixels, each neuron of the distortion classifier is connected to the local perception field of 20 × 20 pixels, and the weight data corresponding to each neuron is 32 × 24.
In the embodiment of the present invention, the convolutional neural network may include three convolutional kernels, where the three convolutional kernels respectively extract three image features, namely, an image size, an inclination left inner angle, and a pincushion distortion radian, corresponding to a sample image included in the training sample image set, and train a distortion classifier through the extracted image features. The structure of the convolutional neural network and the process of training the convolutional neural network are as follows:
inputting: 640 x 480 size sample image; the first layer of convolution: a convolution kernel of size 20 x 20, the convolution kernel to learn the oblique left inner angle; first-layer pooling: 2 x 2 kernels for performing an aggregation operation on the first layer of convolution learned oblique left interior angles; second layer convolution: 5 × 5 convolution kernel for learning image size; second-layer pooling: 2 x 2 kernels for performing an aggregation operation on the convolution-learned image sizes of the second layer; and a third layer of convolution: 2 x 2 convolution kernel for learning pincushion distortion radians; and (3) third-layer pooling: 2 x 2 kernel, the layer for performing an aggregation operation on the third layer of convolution-learned pincushion distortion radians; an output layer: 100 neurons.
After the distortion classifier is trained in advance through the operations of the above steps S1-S5, the distortion classifier is used to perform distortion calibration on the image capturing device through the operations of the steps 101-103, as shown in fig. 2.
Step 101: and acquiring a distorted image of the image acquisition equipment.
Acquiring an original image through image acquisition equipment to be calibrated, and acquiring the original image from the image acquisition equipment, wherein the original image is an acquired distorted image of the image acquisition equipment. Wherein, the image collecting device may be a device with a fixed light path and a fixed collecting direction, for example: fingerprint gathering instrument.
Step 102: classification data for the distorted image is determined by a pre-trained distortion classifier.
The classification data is a classification serial number, and may be a number or a letter, such as 0, 1 or 2, and further, for example, a, b or c. In the embodiment of the invention, when the distortion category is preset, classification data corresponding to each distortion category is also set.
Image features of the distorted image are identified by a pre-trained distortion classifier, the image features including image size, number of left inner corners of tilt, and pincushion distortion radians. Because the pre-trained distortion classifier learns the image characteristics corresponding to each distortion category, the distortion category to which the distorted image belongs can be determined through the distortion classifier according to the identified image size, the left inner angle degree of inclination and the pincushion distortion radian. And determining the classification data corresponding to the determined distortion class as the classification data corresponding to the distorted image.
Step 103: burning the classified data into the image acquisition equipment, and automatically correcting distortion of the distorted image according to the classified data.
Because the classified data is a serial number, the data volume is very small, the classified data can be burned into the image acquisition equipment in a short time, the calibration time of the image acquisition equipment is shortened, and the calibration efficiency is improved.
After the classification data are burned into the image acquisition equipment, the automatic distortion correction processing of the distortion image acquired by the image acquisition equipment can be realized according to the classification data. A set of standard distortion parameters is pre-constructed prior to the automatic distortion correction process. Specifically, a standard distortion image corresponding to each distortion category may be acquired. For each distortion category, a grid is divided on a standard distortion image corresponding to the distortion category, intersection point coordinates on the grid are obtained, and the obtained intersection point coordinates are determined as standard distortion parameters corresponding to the distortion category. The standard distortion parameters corresponding to each distortion category are obtained according to the method, the operation of pre-constructing the standard distortion parameters is completed, and then the distortion image acquired by the image acquisition equipment can be automatically corrected according to the classification data burnt in the image acquisition equipment and the constructed standard distortion parameters.
Classification data is first read from the image capture device. And according to the classification data, distortion correction processing is carried out on the distorted image acquired by the image acquisition equipment. Specifically, the distortion correction processing is performed on the distorted image acquired by the image acquisition device in the following way, and the method comprises the following steps:
acquiring distortion data and standard distortion parameters corresponding to the classification data; calculating distortion parameters corresponding to the image acquisition equipment through an image affine algorithm according to the distortion data and the standard distortion parameters; and correcting the image acquired by the image acquisition equipment according to the distortion parameter. The image affine algorithm may include an image scaling algorithm, an image shifting algorithm, and an image warping algorithm. The image scaling algorithm, the image shifting algorithm and the image warping algorithm are commonly used by those skilled in the art, and are not described herein again.
When distortion correction processing is carried out on a distorted image according to distortion parameters, distortion correction is carried out by adopting a proportional mapping algorithm, grids are divided on the distorted image acquired by image acquisition equipment, intersection point coordinates on the divided grids are acquired, and the intersection point coordinates are subjected to scaling processing through the proportional mapping algorithm according to the distortion parameters of the image acquisition equipment, so that an image after distortion correction is obtained. As shown in fig. 3, a is an image acquired by the image acquisition apparatus and having distortion, and b is an image after the distortion processing is corrected in the above manner.
In order to facilitate understanding of the image distortion correction process according to the classification data, the image distortion correction process will be briefly described below with reference to the drawings. As shown in fig. 4, a 1: and initializing the image acquisition equipment. A2: and reading the classification data from the image acquisition equipment. A3: and calculating distortion parameters according to the classification data. A4: the image capture device is turned on. A5: and reading the distorted image from the image acquisition equipment, and performing distortion correction processing on the distorted image according to the distortion parameters. A6: and outputting the corrected image.
In the embodiment of the invention, the pre-trained distortion classifier can simultaneously process a plurality of distorted images and output classification data corresponding to the plurality of distorted images in a classification manner. Therefore, the embodiment of the invention can simultaneously carry out distortion calibration on a plurality of image acquisition devices, realize batch calibration of the image acquisition devices and greatly improve the calibration efficiency of the image acquisition devices.
According to the embodiment of the invention, the deep learning technology is firstly applied to the calibration work of the image acquisition equipment, the workload of the calibration work of the current image acquisition equipment is reduced, the automatic calibration is realized, the human input and the time input are greatly reduced, and the batch calibration of the image acquisition equipment is realized.
In the embodiment of the invention, a distorted image of an image acquisition device is acquired; determining classification data of a distorted image through a pre-trained distortion classifier; burning the classification data into image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data. The classification data corresponding to the distorted image is determined through the distortion classifier, only the classification data is burnt into the image acquisition equipment, the classification data is the serial number corresponding to the distortion class to which the distorted image belongs, the data volume is small, the time spent on burning the classification data is short, the calibration time is shortened, and the calibration efficiency is improved. And the pre-trained distortion classifier can simultaneously carry out distortion classification on multi-frame distortion images, so that parallel calibration on a plurality of image acquisition devices is realized, batch calibration of the image acquisition devices is realized, and the calibration efficiency of the image acquisition devices is greatly improved.
Example 2
As shown in fig. 5, a calibration apparatus for an image capturing device, the apparatus being configured to execute the calibration method for the image capturing device provided in embodiment 1, the apparatus specifically includes:
the acquisition module 20 is used for acquiring a distorted image of the image acquisition equipment;
a determining module 21, configured to determine classification data of a distorted image through a pre-trained distortion classifier;
the burning module 22 is configured to burn the classification data into the image acquisition device, and automatically correct distortion of the distorted image according to the classification data.
The determining module 21 may include:
the identifying unit is used for identifying the image characteristics of the distorted image through a pre-trained distortion classifier, wherein the image characteristics comprise the image size, the left inner angle degree of inclination and the pincushion distortion radian;
the determining unit is used for determining the distortion category to which the distorted image belongs through the distortion classifier according to the image characteristics; and determining the classification data corresponding to the distortion classification as the classification data corresponding to the distorted image.
In an embodiment of the present invention, the apparatus may further include:
the classifier training module is used for acquiring a training sample image set and a test sample image set; carrying out classification training on the convolutional neural network according to the sample images included in the training sample image set; when the training of the convolutional neural network meets a preset condition, carrying out distortion classification on the test images included in the test sample image set through the trained convolutional neural network; and obtaining a classification error corresponding to the trained convolutional neural network, and determining the trained convolutional neural network as a distortion classifier when the classification error is smaller than a preset threshold value.
The burning module is used for reading the classified data from the image acquisition equipment and acquiring distortion data and standard distortion parameters corresponding to the classified data; according to the distortion data and the standard distortion parameters, calculating distortion parameters corresponding to the image acquisition equipment through an image affine algorithm; and performing distortion correction processing on the image acquired by the image acquisition equipment according to the distortion parameters.
In an embodiment of the present invention, the apparatus may further include:
and the standard distortion parameter acquisition module is used for acquiring a standard distortion image corresponding to the distortion category, dividing the standard distortion image into grids, acquiring the coordinates of the intersection points of the grids divided on the standard distortion image, and determining the coordinates of the intersection points as the standard distortion parameters corresponding to the distortion category.
In the embodiment of the invention, a distorted image of an image acquisition device is acquired; determining classification data of a distorted image through a pre-trained distortion classifier; burning the classification data into image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data. The classification data corresponding to the distorted image is determined through the distortion classifier, only the classification data is burnt into the image acquisition equipment, the classification data is the serial number corresponding to the distortion class to which the distorted image belongs, the data volume is small, the time spent on burning the classification data is short, the calibration time is shortened, and the calibration efficiency is improved. And the pre-trained distortion classifier can simultaneously carry out distortion classification on multi-frame distortion images, so that parallel calibration on a plurality of image acquisition devices is realized, batch calibration of the image acquisition devices is realized, and the calibration efficiency of the image acquisition devices is greatly improved.
The calibration device of the image acquisition equipment provided by the embodiment of the invention can be specific hardware on the equipment or software or firmware installed on the equipment and the like. The device provided by the embodiment of the present invention has the same implementation principle and technical effect as the method embodiments, and for the sake of brief description, reference may be made to the corresponding contents in the method embodiments without reference to the device embodiments. It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the foregoing systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments provided by the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. The storage medium can be read and executed by a computer, and the effect of the scheme described in the embodiment of the specification is achieved. Accordingly, the present invention also provides a computer readable storage medium for calibrating an image acquisition device, comprising a processor and a memory for storing processor-executable instructions, which when executed by the processor implement steps comprising:
acquiring a distorted image of image acquisition equipment;
determining classification data of the distorted image through a pre-trained distortion classifier;
burning the classification data into the image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data.
Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It should be noted that the description of the above-mentioned computer-readable storage medium according to the method embodiment may also include other embodiments. The specific implementation manner may refer to the description of the related method embodiment, and is not described in detail herein.
The computer-readable storage medium for calibrating the image acquisition device by updating the feature template in iris recognition provided by the embodiment determines the classification data corresponding to the distorted image through the distortion classifier, only burns the classification data into the image acquisition device, the classification data is the serial number corresponding to the distortion class to which the distorted image belongs, the data volume is small, the time spent in burning the classification data is short, the calibration time is shortened, and the calibration efficiency is improved. And the pre-trained distortion classifier can simultaneously carry out distortion classification on multi-frame distortion images, so that parallel calibration on a plurality of image acquisition devices is realized, batch calibration of the image acquisition devices is realized, and the calibration efficiency of the image acquisition devices is greatly improved.
The invention also provides a system for calibrating the image acquisition equipment, which can be a single computer, and can also comprise an actual operation device using one or more methods or devices of one or more embodiments of the specification, and the like. The system for calibrating an image acquisition device may comprise at least one processor and a memory storing computer-executable instructions, which when executed by the processor, implement the steps of the method of any one or more of the above embodiments.
It should be noted that the above-mentioned device may also include other implementation manners according to the description of the method or apparatus embodiment, and specific implementation manners may refer to the description of the related method embodiment, which is not described in detail herein.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus once an item is defined in one figure, it need not be further defined and explained in subsequent figures, and moreover, the terms "first", "second", "third", etc. are used merely to distinguish one description from another and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the present invention in its spirit and scope. Are intended to be covered by the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (8)
1. A calibration method of an image acquisition device, the method comprising:
acquiring a distorted image of image acquisition equipment;
determining classification data of the distorted image through a pre-trained distortion classifier;
burning the classification data into the image acquisition equipment, and automatically correcting distortion of the distorted image according to the classification data;
wherein the automatically correcting the distorted image according to the classification data comprises:
reading classification data from the image acquisition device;
distortion data and standard distortion parameters corresponding to the classification data are obtained;
calculating distortion parameters corresponding to the image acquisition equipment through an image affine algorithm according to the distortion data and the standard distortion parameters;
distortion correction processing is carried out on the distorted image acquired by the image acquisition equipment according to the distortion parameters;
before the automatic distortion correction is performed on the distorted image according to the classification data, the method further comprises the following steps:
and acquiring a standard distortion image corresponding to the classification data, dividing the standard distortion image into grids, acquiring intersection coordinates of the grids divided on the standard distortion image, and determining the intersection coordinates as standard distortion parameters corresponding to the classification data.
2. The method of claim 1, wherein determining classification data for the distorted image by a pre-trained distortion classifier comprises:
identifying image features of the distorted image through a pre-trained distortion classifier, wherein the image features comprise an image size, a left internal angle degree of inclination and a pincushion distortion radian;
determining a distortion category to which the distorted image belongs through the distortion classifier according to the image features;
and determining the classification data corresponding to the distortion category as the classification data corresponding to the distorted image.
3. The method of claim 1, wherein prior to obtaining classification data for the distorted image by the pre-trained distortion classifier, further comprising:
acquiring a training sample image set and a test sample image set;
carrying out classification training on the convolutional neural network according to the sample images included in the training sample image set;
when the training of the convolutional neural network meets a preset condition, carrying out distortion classification on the test images included in the test sample image set through the trained convolutional neural network;
and obtaining a classification error corresponding to the trained convolutional neural network, and determining the trained convolutional neural network as a distortion classifier when the classification error is smaller than a preset threshold value.
4. A calibration device of an image acquisition device, the device comprising:
the acquisition module is used for acquiring a distorted image of the image acquisition equipment;
the determining module is used for determining classification data of the distorted image through a pre-trained distortion classifier;
the burning module is used for burning the classification data into the image acquisition equipment and automatically correcting distortion of the distorted image according to the classification data;
the burning module is used for reading the classified data from the image acquisition equipment and acquiring distortion data and standard distortion parameters corresponding to the classified data; according to the distortion data and the standard distortion parameters, calculating distortion parameters corresponding to the image acquisition equipment through an image affine algorithm; correcting the distortion of the image acquired by the image acquisition equipment according to the distortion parameter;
the calibration device of the image acquisition equipment further comprises:
and the standard distortion parameter acquisition module is used for acquiring a standard distortion image corresponding to the classification data, dividing the standard distortion image into grids, acquiring the coordinates of the intersection points of the grids divided on the standard distortion image, and determining the coordinates of the intersection points as the standard distortion parameters corresponding to the classification data.
5. The apparatus of claim 4, wherein the determining module comprises:
the identifying unit is used for identifying the image characteristics of the distorted image through a pre-trained distortion classifier, wherein the image characteristics comprise the image size, the left inner angle degree of inclination and the pincushion distortion radian;
the determining unit is used for determining a distortion category to which the distorted image belongs through the distortion classifier according to the image characteristics; and determining the classification data corresponding to the distortion category as the classification data corresponding to the distorted image.
6. The apparatus of claim 4, further comprising:
the classifier training module is used for acquiring a training sample image set and a test sample image set; carrying out classification training on the convolutional neural network according to the sample images included in the training sample image set; when the training of the convolutional neural network meets a preset condition, carrying out distortion classification on the test images included in the test sample image set through the trained convolutional neural network; and obtaining a classification error corresponding to the trained convolutional neural network, and determining the trained convolutional neural network as a distortion classifier when the classification error is smaller than a preset threshold value.
7. A computer-readable storage medium for calibrating an image acquisition device, comprising a processor and a memory for storing processor-executable instructions, which when executed by the processor, implement the steps of the method of calibrating an image acquisition device according to any one of claims 1-3.
8. A system for calibrating an image acquisition device, comprising at least one processor and a memory storing computer-executable instructions, which when executed by the processor implement the steps of the method of calibrating an image acquisition device according to any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810615799.3A CN110610522B (en) | 2018-06-14 | 2018-06-14 | Calibration method and device of image acquisition equipment, readable storage medium and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810615799.3A CN110610522B (en) | 2018-06-14 | 2018-06-14 | Calibration method and device of image acquisition equipment, readable storage medium and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110610522A CN110610522A (en) | 2019-12-24 |
| CN110610522B true CN110610522B (en) | 2022-02-08 |
Family
ID=68887741
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810615799.3A Active CN110610522B (en) | 2018-06-14 | 2018-06-14 | Calibration method and device of image acquisition equipment, readable storage medium and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110610522B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20170059040A (en) * | 2015-11-19 | 2017-05-30 | 전자부품연구원 | Optimal mode decision unit of video encoder and video encoding method using the optimal mode decision |
| CN107248178A (en) * | 2017-06-08 | 2017-10-13 | 上海赫千电子科技有限公司 | A kind of fisheye camera scaling method based on distortion parameter |
-
2018
- 2018-06-14 CN CN201810615799.3A patent/CN110610522B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20170059040A (en) * | 2015-11-19 | 2017-05-30 | 전자부품연구원 | Optimal mode decision unit of video encoder and video encoding method using the optimal mode decision |
| CN107248178A (en) * | 2017-06-08 | 2017-10-13 | 上海赫千电子科技有限公司 | A kind of fisheye camera scaling method based on distortion parameter |
Non-Patent Citations (2)
| Title |
|---|
| Radial Lens Distortion Correction Using Convolutional Neural Networks Trained with Synthesized Images;Jiangpeng Rong et.al;《Computer Vision –ACCV 2016》;20170311;第1-3、5节 * |
| 基于卷积神经网络的图书页面检索方法;刘乐元;《华中科技大学学报(自然科学版)》;20171123;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110610522A (en) | 2019-12-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109583445B (en) | Text image correction processing method, device, equipment and storage medium | |
| CN110060237B (en) | Fault detection method, device, equipment and system | |
| CN108764247B (en) | Dense connection-based deep learning object detection method and device | |
| CN109376631B (en) | Loop detection method and device based on neural network | |
| CN106503724A (en) | Grader generating means, defective/zero defect determining device and method | |
| CN109740617A (en) | A kind of image detecting method and device | |
| US10861187B2 (en) | Method of processing object detection data | |
| CN111401472B (en) | Infrared target classification method and device based on deep convolutional neural network | |
| CN110688883A (en) | Vehicle and pedestrian detection method and device | |
| JP7277341B2 (en) | Curved object segmentation with noise antecedents | |
| CN113269752A (en) | Image detection method, device terminal equipment and storage medium | |
| CN112381177A (en) | Dial digital character recognition method and system based on deep learning | |
| CN110007764B (en) | Gesture skeleton recognition method, device and system and storage medium | |
| CN109697442A (en) | A kind of training method and device of character recognition model | |
| CN112784494B (en) | Training method of false positive recognition model, target recognition method and device | |
| JP7277342B2 (en) | Curved object segmentation using geometric antecedents | |
| CN110610522B (en) | Calibration method and device of image acquisition equipment, readable storage medium and system | |
| CN111553250B (en) | Accurate facial paralysis degree evaluation method and device based on face characteristic points | |
| CN110765852B (en) | Method and device for acquiring face direction in image | |
| CN111832550A (en) | Data set manufacturing method and device, electronic equipment and storage medium | |
| CN116524508A (en) | Correction method and device for table type image, storage medium and computer equipment | |
| CN113537158B (en) | Image target detection method, device, equipment and storage medium | |
| CN115019377A (en) | Eye state judgment method, eye state judgment device and electronic equipment | |
| CN113591601B (en) | Method and device for identifying hyphae in cornea confocal image | |
| CN111553249B (en) | H-B grading-based accurate facial paralysis degree evaluation method and device under CV |
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 | ||
| CP02 | Change in the address of a patent holder |
Address after: 071800 Beijing Tianjin talent home (Xincheng community), West District, Xiongxian Economic Development Zone, Baoding City, Hebei Province Patentee after: BEIJING EYECOOL TECHNOLOGY Co.,Ltd. Address before: 100085 804, 8th floor, building 1, yard 1, shangdishi street, Haidian District, Beijing Patentee before: BEIJING EYECOOL TECHNOLOGY Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Calibration method, device, readable storage medium, and system for image acquisition equipment Effective date of registration: 20230510 Granted publication date: 20220208 Pledgee: Bank of China Limited Hebei xiong'an branch Pledgor: BEIJING EYECOOL TECHNOLOGY Co.,Ltd. Registration number: Y2023980040255 |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right | ||
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Granted publication date: 20220208 Pledgee: Bank of China Limited Hebei xiong'an branch Pledgor: BEIJING EYECOOL TECHNOLOGY Co.,Ltd. Registration number: Y2023980040255 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Calibration method, device, readable storage medium, and system for image acquisition equipment Granted publication date: 20220208 Pledgee: Bank of China Limited Hebei xiong'an branch Pledgor: BEIJING EYECOOL TECHNOLOGY Co.,Ltd. Registration number: Y2024110000186 |