KR20180072020A - Method, apparatus and computer program stored in computer readable medium for correction of image data - Google Patents

Abstract

According to an embodiment of the present invention, disclosed is a method for correcting target image data, which comprises the steps of: generating reference image data on the basis of a plurality of normal image data without defects; detecting one or more reference image feature points from the reference image data; detecting one or more target image feature points from target image data; determining a target image estimated feature point and a reference image estimated feature point for each of an undetected feature point and an unmatched feature point from the reference image data and the target image data, on the basis of at least a part of an algorithm for estimating feature points; matching the reference image data and the target image data on the basis of at least the reference image feature points, the target image feature points, and the estimated feature points; and recovering at least a part of the target image data lost by illumination.

Description

[0001] METHOD, APPARATUS AND COMPUTER PROGRAM STORED IN COMPUTER READABLE MEDIUM FOR CORRECTION OF IMAGE DATA [0002]

The present invention relates to a method of correcting image data, and more particularly, to a method of correcting image data for determining the state of image data.

In the process, targets to be inspected are imaged and checked for defects. On the other hand, there is a limit in detecting and analyzing image data by a human, and a method for determining the state of image data with high accuracy is required in the art. A neural network may be used to implement a method for determining the state of image data with high accuracy. U.S. Patent No. 7,698,239 shows an example of such a neural network.

On the other hand, in the image data description area, noise refers to a signal in an undesired form added to the pixel value of the image data. Geometric distortion due to lens error, and uneven illumination. Such geometric distortion phenomenon and irregularity of illumination may cause a problem of insubstantiation in the judgment system as the machine vision system for judging the state of image data becomes more precise.

Accordingly, there is a need in the art to solve the geometric distortion and illumination unevenness of the image data.

The present invention is devised in correspondence with the background art described above, and it is an object of the present invention to provide a method of correcting image data.

An object of the present invention is to provide state information of image data through a method of correcting image data.

A method for correcting target image data according to embodiments of the present disclosure to solve the above-described problems is disclosed. The method includes the steps of: generating reference image data based on a plurality of normal image data free from defects; detecting one or more reference image feature points from the reference image data; Detecting one or more target image feature points from the target image data; Determining a target image feature point and a reference image feature point for each of the reference image data and non-detected feature points and unmatched feature points from the target image based at least in part on an algorithm for feature point estimation, Matching the reference image data and the target image data based at least on the target image feature point and the estimated feature point; And restoring data of at least a portion of the target image data lost by the illumination; . ≪ / RTI >

Alternatively, the step of detecting the one or more reference image feature points and the one or more target image feature points may include: dividing the reference image data and the target image data into one or more regions having a predetermined size, And detecting a minutiae point.

Alternatively, the predetermined size may be a size that is set based on at least one of the image resolution, size, color mode, and aspect ratio of the reference image data and the target image data.

Alternatively, the step of detecting feature points from the reference image data may include: detecting a feature point candidate group from the reference image data based at least on a first algorithm for feature point detection; Detecting a feature point necessary for matching with the target image data from the feature point candidate group based at least on a second algorithm for feature point detection; As shown in FIG.

Alternatively, the step of detecting one or more feature points from the target image data may comprise: determining, as feature points, position coordinates on the target image data corresponding to position coordinates for each of the feature points detected from the reference image data, Detecting a feature point candidate group from the image data; Detecting a feature point necessary for matching with the reference image data from the feature point candidate group based at least on a second algorithm for feature point detection; As shown in FIG.

Alternatively, the first algorithm for feature point detection may be at least one of a Harris Corner Detector algorithm, a Shi-Tomasi Corner Detector algorithm, and a SIFT_DoG algorithm.

Alternatively, a second algorithm for the feature point detection may comprise a normalized cross-correlation between the reference image data and second image data associated with the reference image data and between the target image data and second image data associated with the target image data (NCC) result data may be a Super Resoulution Feature Matching algorithm that analyzes the data in units of subpixels.

Alternatively, the second image data may be generated by performing at least one of deformation and loss on the image data.

Alternatively, the second image data may be generated by applying a Gaussian Blur filter and a Gaussian noise to the image data.

Alternatively, the algorithm for feature point estimation may be configured to include at least a conversion model equation corresponding to the number of feature points, based on at least the number of feature points and the distance connecting each feature point.

Alternatively, the transformation model expression may be an Offset transformation when the number of the feature points is 1, a Similarity transformation when the number of the feature points is 2, an Affine transformation when the number of the feature points is 3, Perspective transformation can be determined.

Alternatively, the step of matching the reference image data and the target image data comprises the steps of: transforming the target image data; And normalizing the size of the target image data.

Alternatively, the method may further include determining status information of the target image data using the network function learned for the reference image data.

Alternatively, the status information of the target image data may be determined based on status information for each of the one or more areas divided into a predetermined size.

An apparatus for correcting target image data in accordance with embodiments of the present disclosure is disclosed. The apparatus includes: a reference image generated based on a plurality of normal image data free from defects; an image data acquiring unit acquiring the target image data; A first detector for detecting one or more feature points from the reference image data; A second detecting unit detecting at least one feature point from the target image data; An estimated feature point determiner for determining an estimated feature point for each of the feature point and the unmatched feature point that are not detected from the reference image data and the target image data based at least on an algorithm for feature point estimation; An image data matching unit for matching the reference image data and the target image data based at least on the feature points and the estimated feature points; And a data restoring unit for restoring at least a part of the data of the target image data lost by illumination. . ≪ / RTI >

Alternatively, the image data obtaining section may include: a photographing module for photographing a target; A first acquiring module for acquiring a plurality of normal image data without defects; And a reference image generation module that generates the reference image data based on the plurality of normal image data.

Alternatively, the image data obtaining section may include: a photographing module for photographing a target; And a second acquisition module for acquiring the reference image generated based on the plurality of normal image data without defects.

Alternatively, the image data obtaining section may include a third obtaining module for obtaining the reference image data generated based on the plurality of normal image data without defects and the target image data taken by the photographing module have.

Alternatively, the apparatus may further include a status information determination unit for determining status information of the target image data using the network function learned for the reference image data.

A computer program stored on a computer readable storage medium is disclosed that includes a plurality of instructions executed by one or more processors of an apparatus for correcting target image data in accordance with embodiments of the present disclosure. The computer program comprising instructions for: obtaining a reference image and a target image data generated based on a plurality of normal image data that are free from defects; Detecting one or more feature points from the reference image data; Detecting one or more feature points from the target image data; Determining an estimated feature point for each of the feature point and non-matched feature point that are not detected from the reference image data and the target image data, based at least on an algorithm for feature point estimation; Instructions for matching the reference image data and the target image data based at least on the feature points and the estimated feature points; And restoring at least a portion of the data of the target image data lost by the illumination.

The present invention can provide a method of correcting image data.

The present invention can provide status information of image data.

In order that the features of the above-mentioned subject matter of the present invention will be understood in detail and with reference to the following embodiments in more detail, some of the embodiments are shown in the accompanying drawings. In addition, like reference numerals in the drawings are intended to refer to the same or similar functions throughout the several views. It should be understood, however, that the appended drawings illustrate only typical exemplary embodiments of the present invention and are not to be considered limiting of its scope, and that other embodiments having the same effect may be fully recognized Please note.
1 is a diagram illustrating reference image data and target image data according to embodiments of the present disclosure.
2 is a flow chart illustrating a method of correcting target image data according to embodiments of the present disclosure.
Figure 3 is a diagram illustrating an area of reference image data according to embodiments of the present disclosure and a method for detecting feature points from any area of the reference image data.
4 is a diagram illustrating the concept that second image data associated with reference image data is generated from reference image data according to embodiments of the present disclosure.
5 illustrates a diagram for describing a second algorithm for feature point detection according to embodiments of the present disclosure.
FIG. 6 is a diagram showing non-detected and / or unmatched feature points and their estimated feature points according to embodiments of the present disclosure; FIG.
7 is a diagram illustrating a concept of matching target image data according to embodiments of the present disclosure.
8 is a block diagram showing the configuration of a correction apparatus for target image data according to the embodiments of the present disclosure.
9 is a first embodiment of an image data obtaining section according to the embodiments of the present disclosure.
10 is a second embodiment of the image data obtaining section according to the embodiments of the present disclosure.
11 is a third embodiment of the image data obtaining section according to the embodiments of the present disclosure.
12 is a conceptual diagram illustrating an example of a neural network according to embodiments of the present disclosure.

Various embodiments are now described with reference to the drawings. In this specification, various explanations are given in order to provide an understanding of the present disclosure. It will be apparent, however, that such embodiments may be practiced without these specific details.

In addition, the term "or" is intended to mean " exclusive or " That is, it is intended to mean one of the natural inclusive substitutions "X uses A or B ", unless otherwise specified or unclear in context. That is, X uses A; X uses B; Or when X uses both A and B, "X uses A or B" can be applied to either of these cases. It should also be understood that the term "and / or" as used herein refers to and includes all possible combinations of one or more of the listed related items.

It is also to be understood that the term " comprises "and / or" comprising " means that the feature and / or component is present, but does not exclude the presence or addition of one or more other features, components and / It should be understood that it does not. Also, unless the context clearly dictates otherwise or to the contrary, the singular forms in this specification and claims should generally be construed to mean "one or more. &Quot;

The description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features presented herein.

1 is a diagram illustrating reference image data and target image data according to embodiments of the present disclosure.

As used herein, the term "normal image data" refers to image data determined to be free of defects. The normal image data can be determined by the judgment of the user. Alternatively, the network function learned for the reference image data may refer to images for which the status information of the target image data is determined as " normal ". Reference image data may be generated based on a plurality of normal image data according to embodiments of the present disclosure.

As used herein, the term "reference image data" refers to image data generated based on a plurality of normal image data that are free from defects as shown in Fig. 1 (a). This reference image data can be referred to for correcting the target image data.

Reference image data according to embodiments of the present disclosure may be generated by an apparatus 100 for correcting target image data, which will be described below. Alternatively, the reference image data according to embodiments of the present disclosure may be received from the outside via the network portion 180. In another example, reference image data may be input via an input (e.g., a USB input, etc.). Such reference image data may be partitioned into image patches of a size that can be input to the network function. Or the reference image data itself can be preset to a size that can be input to the network function.

As used herein, the term "target image data" refers to an image in which an object (i.e., a target) for which a defect status is to be determined, as shown in Fig. According to embodiments of the present disclosure, target image data may be obtained from apparatus 100 that corrects target image data. For example, the target image data may be obtained from the photographing module 111 of the apparatus 100 for correcting the target image data as shown in Fig. 1 (b).

In order to determine the state of the target image data (for example, the presence or absence of a defect, the type of a defect, etc.), the target image data needs to be physically similar to the reference image data. On the other hand, when acquiring the target image data, the geometric distortion (e.g., distortion caused by the lens module, mechanical defects of the imaging module 111, and target motion, etc.) Loss such as crushing of data, etc.) occurs. As an example, a target moving on a conveyor belt may experience slippage, movement, rotation, vibration, etc., and these factors are displayed in many geometric distortion states within the acquired target image data. Moreover, when acquiring the target image data, a distortion phenomenon due to non-uniformity of illumination may also occur.

This geometric distortion phenomenon and uneven illumination can be fatal as the machine vision system for determining the state of the image data is more precise. For example, the above-described distortion phenomenon may not be a serious problem in an inaccurate machine vision system with a pixel resolution of 0.5 mm or more, but a precision system with a pixel resolution of 0.1 mm or less, Can be a very deadly problem that can not be tested.

The target image data correction method according to the embodiments of the present disclosure described below allows the target image data to be corrected. For example, the target image data correcting method according to the embodiments of the present disclosure is characterized in that when the target image data is acquired using the photographing module 111, 1) to correct the photographed target image data.

The target image data can be corrected according to the method of correcting the target image data described below. Accordingly, the determination of the state of the target image data can be facilitated. In addition, the accuracy of the determination of the state of the target image data can be improved.

As used herein, the terms "reference image data" and "reference image" are often used interchangeably. As used herein, the terms "target image data" and "target image" are often used interchangeably. Further, the terms "reference image data" and "target image data" used herein may be collectively referred to as "image data ". In addition, the term "apparatus 100 and apparatus 100 for correcting target image data" as used herein may often be used interchangeably.

Hereinafter, a method for correcting target image data according to embodiments of the present disclosure will be described in detail with reference to FIGS. 2 to 7. FIG.

2 is a flow chart illustrating a method of correcting target image data according to embodiments of the present disclosure.

The method of correcting target image data according to embodiments of the present disclosure described below will be described with reference to FIG. 1 and to a target image data correction apparatus 100 according to embodiments of the present disclosure, But the scope of the rights of the present disclosure is not limited thereto. A method of correcting target image data according to embodiments of the present disclosure may be performed by a computer-readable storage medium including a plurality of instructions for performing the method, and the scope of rights of the present disclosure is not limited thereto.

According to embodiments of the present disclosure, a method of correcting target image data includes generating reference image data (S110), detecting one or more feature points from reference image data (S120), detecting one or more feature points from the target image data (S140) determining an estimated feature point for each of the non-detected feature points and unmatched feature points, and matching the reference image data and the target image data (S150). In addition, a method of correcting target image data according to embodiments of the present disclosure may include reconstructing (S160) at least a portion of the data of the target image data lost by illumination (S160) and a learned network function (S170) the status information of the target image data. The steps as described above are exemplary steps in accordance with the embodiments of the present disclosure, and the scope of rights of the present disclosure is not limited thereto.

A method of correcting target image data according to embodiments of the present disclosure may generate reference image data (S110). The reference image data can be generated based on a plurality of normal image data without defects, as described above with reference to Fig. Prior to step S110, one or more normal image data for generating reference image data may be input.

Detecting one or more feature points from the reference image data (S120) may be performed according to embodiments of the present disclosure. Feature points according to embodiments of the present disclosure may include corners, edges, contours, line intersections, and the like. That is, the feature point refers to representative pixels capable of expressing the visual characteristic of the objects in the image.

According to embodiments of the present disclosure, a feature point candidate group may be detected from reference image data based at least on a first algorithm for feature point detection (S121). The first algorithm for feature point detection according to embodiments of the present disclosure may be at least one of a Harris Corner Detector algorithm, a Shi-Tomasi Corner Detector algorithm, and a SIFT_DoG algorithm. In the embodiments of the present disclosure, the feature point extraction algorithm is not particularly limited, and various feature point extraction algorithms can be used. For example, at least one of the SUSAN algorithm, the SURF algorithm, and the MSER algorithm may be employed as the first algorithm for feature point detection.

When a feature point candidate group is detected through step S121, a feature point that can be actually matched among the feature point candidate groups can be determined. According to embodiments of the present disclosure, feature points necessary for matching with target image data can be detected based at least on a second algorithm for feature point detection (S123).

According to embodiments of the present disclosure, a second algorithm for feature point detection includes interpolating normalized cross correlation (NCC) result data of second image data associated with reference image data and reference image data on a subpixel basis Which may be a Super Resoulution Feature Matching algorithm.

Second image data associated with reference image data in accordance with embodiments of the present disclosure may be generated by performing at least one of transformation and loss on the image data. More specifically, the second image data associated with the reference image data may be generated by applying a Gaussian Blur filter and a Gaussian noise to the reference image data, and the scope of the rights of the present disclosure is not limited thereto Do not.

Figure 3 is a diagram illustrating an area of reference image data according to embodiments of the present disclosure and a method for detecting feature points from any area of the reference image data.

As shown in the figure, the reference image data is divided into a predetermined size and feature points that can be matched in each individual region can be detected. 3, feature points (K1, K2,..., K9) that can be matched among the feature point candidate groups and the feature point candidate groups detected in each individual region according to the method of correcting the target image data according to the embodiments of the present disclosure Respectively.

4 is a diagram illustrating the concept that second image data associated with reference image data is generated from reference image data according to embodiments of the present disclosure.

Second image data associated with reference image data according to embodiments of the present disclosure may be generated by applying a Gaussian blur filter and Gaussian noise to the reference image data. These Gaussian blur filters and Gaussian noise perform Gaussian blur filter and Gaussian noise using the characteristic of Gaussian distribution. For example, the Gaussian blur filter is applied by calculating the sum of values obtained by multiplying the pixels of the reference image data corresponding to the respective positions by the Gaussian blur filter coefficient. These Gaussian blur filters and Gaussian noise use the two-dimensional Gaussian distribution given below.

For reference, x means the x-coordinate of each pixel value of the reference image data. and y represents the y coordinate of each pixel value of the reference image data. The Gaussian filter according to the embodiments of the present disclosure is a continuous function indicating the Gaussian distribution, but the value of the Gaussian function can be calculated in the case where x and y are integers. In a Gaussian distribution with an average of zero and a standard deviation of b, each pixel value of the reference image data is distributed within a range, and the larger the standard deviation, the wider the range of blurring may occur.

In FIG. 4, reference image data is described as an example, but second image data related to the target image data may also be generated according to the above-described method.

2, a feature point candidate group is detected through step S121, and a feature point capable of actually matching among the feature point candidate groups can be determined. According to embodiments of the present disclosure, feature points necessary for matching with target image data can be detected based at least on a second algorithm for feature point detection (S123).

The second algorithm for detecting the feature points may be a Super Resoulution Feature Matching algorithm that analyzes the normalized cross correlation (NCC) result data of the second image data related to the reference image data and the reference image data in units of subpixels have. Referring to Fig. 5, a second algorithm for feature point detection will be described.

5 illustrates a diagram for describing a second algorithm for feature point detection according to embodiments of the present disclosure.

Figure 5 (a) is image data according to embodiments of the present disclosure, and Figure 5 (b) is second image data relating to the image data according to embodiments of the present disclosure. For example, Figure 5 (a) is reference image data according to embodiments of the present disclosure, and Figure 5 (b) is second image data relating to the image data according to embodiments of the present disclosure.

According to the embodiments of the present disclosure, a first template T ₁ having an arbitrary size can be set around any point C on the reference image data. The size of the first template T ₁ can be preset and stored.

The position corresponding to the arbitrary point C may be set to < RTI ID = 0.0 > a < / RTI > A second template T ₂ having an arbitrary size centered on the arbitrary point C 'may be set. The size of the second template T ₂ may be set to be larger than the size of the first template T ₁ . For example, the size of the second template T ₂ may be set to twice the size of the _{first table} T ₁ .

According to the embodiments of the present disclosure, the normalized cross correlation (NCC) of the reference image data (FIG. 5A) and the second image data related to the reference image data (FIG. 5B) is calculated.

The NCC measures the linear difference in brightness and geometric similarity between two image data in a way that finds a normalized correlation. According to the method of correcting target image data according to embodiments of the present disclosure, second image data related to the reference image data can be generated from reference image data, so that it is necessary to acquire additional image data to calculate the NCC value There is no.

According to the method of correcting the target image data according to the embodiments of the present disclosure, an NCC score map can be generated in which the NCC value between the reference image data and the second image data related to the reference image data is converted into a value of 8-bit brightness . According to embodiments of the present disclosure, the following conditions are satisfied:

1) there is only one peak on the NCC score map that exceeds a predetermined threshold,

2) the calculated value obtained by approximating a data matrix of a predetermined size obtained with respect to the peak in the NCC score to a second-order Gaussian distribution (see Equation 1 in FIG. 4) exceeds a predetermined threshold value, and

3) The maximum value of the outermost data value calculated from the data matrix of a predetermined size obtained centering on the peak in the NCC score is less than or equal to a predetermined threshold value

It is determined that the arbitrary point C in Fig. 5 (a) is determined as a minutia.

Again, returning to Fig. 2, one or more feature points may be detected from the target image data (S130).

The target image data correcting method according to the embodiments of the present disclosure is characterized by determining the position coordinates on the target image data corresponding to the position coordinates for each of the detected minutiae from the reference image data as minutiae points, And detecting a candidate group (S131). The target image data correcting method according to the embodiments of the present disclosure includes a step (S133) of detecting a feature point necessary for matching with the reference image data from the feature point candidate group based at least on a second algorithm for feature point detection can do.

Through steps S120 through S130 according to embodiments of the present disclosure, it is possible to detect and detect representative pixels that can represent objects in the image data, respectively, in the reference image data and the target image data. The step S120 of detecting one or more reference image feature points and the step S130 of detecting one or more target image feature points may divide the reference image data and the target image data into at least one region having a predetermined size, By detecting a feature point within the image.

The predetermined size according to embodiments of the present disclosure may be set based on at least one of the image resolution, size, color mode and aspect ratio of the reference image data and the target image data. In addition, the reference image data and the target image data may be separated into image patches of a size that can be input to the network function. Or the reference image data may be preset to a size that can be input to the network function.

In accordance with embodiments of the present disclosure, estimated feature points for each of an unrecognized feature point and an unmatched feature point may be determined (S140). In more detail, the reference image data and the target image estimating feature point and the reference image estimating feature point for each of the non-detected and unmatched feature points from the target image may be determined based at least in part on the algorithm for feature point estimation.

An algorithm for feature point estimation according to embodiments of the present disclosure may be based at least on the number of feature points and the distance connecting each feature point.

The algorithm for feature point estimation according to the embodiments of the present disclosure is configured to include different transformation model expressions corresponding to the number of feature points. Here, the conversion model expression is determined by Offset conversion when the number of the minutiae is one, Similarity conversion when the number of minutiae is two, Affine conversion when the number of minutiae is three, and Perspective conversion when the number of minutiae is four. In this regard, the following equations are referred to. The description will be made with reference to Fig.

FIG. 6 is a diagram showing non-detected and / or unmatched feature points and their estimated feature points according to embodiments of the present disclosure; FIG.

6A indicates a feature point detected on the image data and / or a matching feature point. A ring-shaped point shown in (b) of FIG. 6 indicates estimated feature points estimated corresponding to feature points and / or non-matched feature points not detected on the image data.

In the first grid G ₁ shown in FIG. 6 (b), three characteristic points are detected. In this case, another feature point can be estimated using the affine transformation model equation.

Two, one, and four feature points are detected in the second grid G ₂ , the third grid G ₃ , and the fourth grid G ₄ shown in FIG. 6B, respectively. That is, according to the embodiments of the present disclosure, the similarity transformation, the offset transformation, and the perspective transformation model equation are used for the second grid G ₂ , the third grid G ₃ , and the fourth grid G ₄ , The detected and / or unmatched feature points can be determined as estimated feature points.

D 'of FIG. 6A and D' of FIG. 6B will be described. In accordance with embodiments of the present disclosure, D means a feature point detected on the reference image data. In accordance with embodiments of the present disclosure, D 'means a feature point that is not detected on the target image data and / or mapped to the reference image data. That is, D 'is a feature point that is not detected or matched. This D 'can be estimated through a feature point estimation algorithm satisfying the following equations (2) and (4).

Here, L ₁ denotes a maximum value of a line segment connecting from the position D to the center of gravity of the region i, and L ₂ denotes a maximum value of a line segment connecting each of the minutiae in the image data (generally, Length). Also, n denotes the number of all the regions constituting the image data. C means the conversion model equation. C is determined by Offset transformation, Similarity transformation, Affine transformation and Perspective transformation according to the number of detected minutiae points, as described above.

6 (b) to 6 (c), the estimated feature point D 'on the target image data satisfies the above equations (2) to (4) The position can be estimated.

Referring again to FIG. 2, the reference image data and the target image data may be matched (S150) based at least on the reference image feature point, the target image feature point, and the estimated feature point.

In operation S150, the same feature points are found among other image data based on the feature points extracted from the image data. The feature points are extracted from the image data to generate descriptors, the feature points are extracted from other image data to generate descriptors, By comparing the descriptors, it can be done by finding the same descriptor and finding the same object. That is, descriptors may be used to search for feature points having the same characteristics among features extracted from different image data. In order to accurately locate feature points having the same characteristics, the descriptor includes various information. Since rotations occur frequently, especially for minutiae, information about rotations can be included in the descriptor for comparison considering rotation characteristics. The foregoing is an exemplary description of an image matching method according to embodiments of the present disclosure, and the scope of rights of the present disclosure is not limited thereto.

By the steps S110 to S150 as described above, correction for the geometric distortion for the target image data can be made. Reference is made to Fig. 7 in this regard.

7 is a diagram illustrating a concept of matching target image data according to embodiments of the present disclosure.

Also, according to embodiments of the present disclosure, at least a portion of the data of the target image data lost by illumination may be restored (S160). This makes it possible to eliminate distortion due to uneven illumination of the target image data.

For example, through step S160, noise can be reduced with respect to the target image data, and color correction, color enhancement, gamma correction, color processing, blur processing, edge enhancement processing, and the like can be performed. The foregoing is merely illustrative of exemplary embodiments of the present disclosure, and the scope of the present disclosure is not limited thereto.

According to embodiments of the present disclosure, the state information of the target image data may be determined using the learned network function for the reference image data (S170). The network function will be described later with reference to FIG.

The status information of the target image data may be determined based on status information for each of the one or more areas divided into a predetermined size. For example, by determining the status information for each individual area, the status information for the entire area of the target image data can be determined. The status information of the target image data may be at least one of, for example, the presence or absence of a defect and the type of a defect, and the scope of rights of the disclosure is not limited thereto.

A method of correcting target image data according to embodiments of the present disclosure as described above can be used to determine the status of target image data. A method of determining target image data according to embodiments of the present disclosure can be used, for example, to classify the status of the target image data to be determined.

It will be appreciated by those skilled in the art that the steps of the flowchart shown in FIG. 2 are not essential and that some steps may be omitted or added as needed.

8 is a block diagram showing the configuration of a correction apparatus for target image data according to the embodiments of the present disclosure.

The method of correcting target image data according to embodiments of the present disclosure, as described above, may be performed in devices with computational capability. For example, the method of correcting the target image data may be performed in a personal computer (PC), a work station, a server computing device, and the like. Alternatively, it may be performed by a separate apparatus for performing the method of correcting the target image data according to the embodiments of the present disclosure. A method for correcting target image data according to embodiments of the present disclosure may be performed in one or more computing devices. For example, at least some of the steps of the method of correcting target image data according to embodiments of the present disclosure may be performed at the client device, and some of the steps may be performed at the server device. In this case, the client device and the server device can be connected to each other via a network to transmit and receive the operation result. Alternatively, a method for correcting target image data according to embodiments of the present disclosure may be performed by distributed computing technology.

The apparatus 100 for correcting target image data according to embodiments of the present disclosure includes an image data obtaining unit 110, a first detecting unit 120, a second detecting unit 130, an estimated minutia determining unit 140, A data recovery unit 160, a status information determination unit 170, a network unit 180, 8 are exemplary and in addition the apparatus 100 may further include a user input, a lighting unit, and the like. In addition, the apparatus 100 may further include an output section. The output may include, for example, at least one of a sound output module (e.g., a speaker) and a video output module (e.g., a display). According to embodiments of the present disclosure, the target image data correction apparatus 100 may display image data acquired or corrected and / or matched by the apparatus 100. This will be described later.

Apparatus 100 in accordance with embodiments of the present disclosure may be constructed solely of a configuration of a portion of the components described above. The device may be used in communication with an external server or may itself comprise a server. In addition, some of the configurations of Fig. 8 described above can be integrated. For example, the first detection unit 120 and the second detection unit 130 may be integrated into a detection unit (not shown). As another example, the configuration of some of the estimated minutia determination unit 140, the image data matching unit 150, the data restoration unit 160, and the status information determination unit 170 may be integrated into a control unit (not shown) The scope of rights of disclosure is not limited to those shown. In embodiments, the state information determiner 170 may be included in and operate on an additional device that is operatively connected to the device 100. In the following, the components of the apparatus 100 according to embodiments of the present disclosure will be described below in turn.

According to embodiments of the present disclosure, the image data acquisition section 110 is configured to acquire a reference image and target image data generated based on a plurality of normal image data that are free from defects.

According to an embodiment of the present disclosure, the image data obtaining unit 110 may obtain image data as an image photographed by the photographing module 111. The image data obtaining unit 110 may obtain image data that can be determined from the image input through the input unit. The image data acquiring unit 110 may acquire image data by dividing the image photographed by the photographing module 111 into image patches of a predetermined size.

The image input to the image data acquisition unit 110 can be input in any format such as jpg, png, tif, psd, Ai, etc., and data can be input without limitation to a moving image, an image file, and the like. This image data obtaining unit 110 will be described with reference to Figs. 9 to 11. Fig.

9 is a first embodiment of an image data obtaining section according to the embodiments of the present disclosure.

According to embodiments of the present disclosure, the image data acquisition unit 110 includes a photographing module 111 for photographing a target, a first acquiring module 113 for acquiring a plurality of normal image data free from defects, And a reference image generation module 113 for generating the reference image based on the image data.

The subject to be photographed here may include various images such as a print image, a product printed with predetermined contents, a fabric, leather, and the like. The photographing object described above is merely an example, and the photographing module 111 can take photographs without limitation on the type of photographing objects. The images photographed by the photographing module 111 may be stored in the memory unit 190 or may be transmitted to the outside through the network connection unit 180. The photographing module 111 may be composed of one camera or may be composed of two or more cameras according to the use environment.

In embodiments, the imaging module 111 may be integrally connected to the correction device 100 of the target image data according to the embodiments of the present disclosure, or may be configured as a detachable module. The shooting module 111 may include various hardware configurations to obtain image data for correcting the target image data.

As an example, the photographing module 111 may operate in conjunction with photographing means (see FIG. 1) such as, for example, a conveyor belt, in order to obtain the data for the photograph at high speed. The image taking means (see FIG. 1) may be a component of the device 100 according to the embodiments of the present disclosure, or may be a component of a separate external device. For example, the photographing module 111 may be configured as a line camera for scanning a photographing subject moving along the conveyor belt. The configuration of the photographing module 111 according to the embodiments of the present disclosure is not limited to the above-described configuration, and any combination for efficiently acquiring image data can be used in an embodiment of the present invention.

In addition, the photographing module 111 can photograph various types of images. The photographing result of the photographing module 111 may include a 3D image, a monochrome image, a GIF image stored in accordance with the passage of time, an image image, an infrared image, an electronic image, and the like. The imaging module 111 may also include a film camera, a digital camera, a microscope, a magnifying glass, an infrared camera, an ultraviolet (UV) camera, an X-ray, a magnetic resonance imaging device, and any image acquisition device. The photographing module 111 may be changed in configuration depending on the photographing object to be photographed.

In addition, the image data obtaining unit 110 according to the embodiments of the present disclosure may include a first obtaining module 113 for obtaining a plurality of defect-free normal image data. May be received from the outside via the first acquisition module 113. [ In such a case, the first acquisition module 113 may operate in connection with the network section 180 and / or the input section (e.g., the USB input section, etc.).

In addition, the image data obtaining unit 110 according to the embodiments of the present disclosure may include a reference image generating module 113 that generates a reference image based on a plurality of normal image data.

10 is a second embodiment of the image data obtaining section according to the embodiments of the present disclosure.

According to embodiments of the present disclosure, the image data acquisition unit 110 includes a photographing module 111 for photographing a target, and a second acquiring module 111 for acquiring the reference image generated based on a plurality of normal image data, (115).

According to embodiments of the present disclosure, the reference image data may be received externally by the second acquisition module 115 from the network, or may be input by an additional device (e.g., USB, etc.).

11 is a third embodiment of the image data obtaining section according to the embodiments of the present disclosure.

According to embodiments of the present disclosure, the image data obtaining unit 110 may include a third acquiring module that acquires the reference image generated based on a plurality of normal image data without defects and the target image data photographed by the photographing module, (117).

Referring again to FIG. 8, the first detector 120 according to embodiments of the present disclosure may detect one or more feature points from the reference image data. The second detection unit 130 according to the embodiments of the present disclosure can detect one or more feature points from the target image data. The estimated minutia determining unit 140 according to the embodiments of the present disclosure determines the estimated minutiae for each of the minutiae not detected and the unmatched minutiae from the reference image data and the target image data based at least on the algorithm for minutiae estimation . The first detection unit 120, the second detection unit 130, and the estimated minutia determination unit 140 described above correspond to the steps S120 to S140 in FIG. 2, respectively, and thus the detailed description thereof will be omitted.

The image data matching unit 150 according to the embodiments of the present disclosure performs image matching in which positional relations of two or more image data are associated with each other and aligned in one coordinate system. That is, the image data matching unit 150 aligns the positional relationship between the reference image data and the target image data in one coordinate system. Since the details have been described in detail at step S150 in FIG. 2, further explanation will be omitted.

The data decompression unit 160 according to the embodiments of the present disclosure can restore at least a part of the data of the target image data lost by the illumination.

According to embodiments of the present disclosure, the apparatus 100 for correcting target image data includes a status information determination unit 170 that determines status information of target image data using the learned network function for reference image data can do.

According to an embodiment of the present disclosure, the network unit 180 may transmit the image acquired by the image data acquisition unit 110 to another device or a server. The network portion 180 may include one or more modules that enable wireless communication between the two devices if the image capture device and the image correction device are separately configured. The network unit 180 may include a transmitting unit and a receiving unit. The network unit 180 may include a wired / wireless Internet module for network connection. WLAN (Wi-Fi), Wibro (Wireless broadband), Wimax (World Interoperability for Microwave Access), HSDPA (High Speed Downlink Packet Access) and the like can be used as wireless Internet technologies. Wired Internet technologies include XDSL (Digital Subscriber Line), FTTH (Fiber to the home), and PLC (Power Line Communication).

In addition, the network unit 180 may transmit data to and receive data from an electronic device including a short distance communication module, which is located relatively close to the apparatus 100, including the short distance communication module. Bluetooth, Radio Frequency Identification (RFID), infrared data association (IrDA), Ultra Wideband (UWB), ZigBee, and the like can be used as a short range communication technology. In one embodiment of the present disclosure, the network portion 180 can sense the connection state of the network and the transmission / reception speed of the network. Data received via the network section 180 may be output via the display, stored through the memory section 190, or transmitted to other electronic devices in the vicinity via the local communication module.

According to one embodiment of the present disclosure, the network portion 180 may receive an image from an external device or server. The received image can be judged or classified according to the status of the image data by the control unit. The control unit for extracting image data from the camera unit and the image capturing unit and determining the status of the image data may be configured as a separate device. The received image may be a digital image, a digital image, compressed data, a RAW image, a data file, or the like, and data that can be converted into an image can be received without limitation. The image received from the outside through the network unit 180 may be used for determining the state of the image data in the control unit or may be used for determining the state of the state determining photograph through the state determination of the image.

The memory unit 190 according to the embodiments of the present disclosure may store a program for operation of the control unit and may store input / output data (for example, reference image data information, minutia information detected from reference image data, A first algorithm for feature point detection, a second algorithm for feature point detection, an algorithm for feature point estimation, state determination information, a message, a still image, a moving image, etc.). The memory unit 190 may store data related to display and sound. The memory unit 190 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (for example, SD or XD memory) (Random Access Memory), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM) A magnetic disk, and / or an optical disk.

The memory unit 190 according to embodiments of the present disclosure may store various network functions and algorithms. The network function will be described later with reference to FIG. The memory unit 190 according to embodiments of the present disclosure may store programs, software, instructions, codes, and the like that cause the apparatus 100 to execute. The memory unit 190 according to the embodiments of the present disclosure may temporarily store image data including a cache and temporarily store programs, commands, and the like necessary for operating the apparatus 100. [

The memory unit 190 according to the embodiments of the present disclosure includes a reference image generated based on a plurality of normal image data without defects and an instruction to acquire the target image data; Detecting one or more feature points from the reference image data; Detecting one or more feature points from the target image data; Determining an estimated feature point for each of the feature point and non-matched feature point that are not detected from the reference image data and the target image data, based at least on an algorithm for feature point estimation; Instructions for matching the reference image data and the target image data based at least on the feature points and the estimated feature points; And restoring at least a portion of the data of the target image data lost by the illumination.

The memory unit 190 according to the embodiments of the present disclosure may store at least a part of images and images photographed in the camera module 111. [ In addition, the memory unit 190 may delete, from the stored photographing data, a photographed image that is not loaded by another component such as a controller or another device after a predetermined time. The memory unit 190 may store the photographed image divided into image patches of various sizes. The configuration of the memory unit 190 is merely an example, and the present disclosure is not limited thereto.

In accordance with embodiments of the present disclosure, apparatus 100 may include an input. The input unit may connect an external device to use the device, or may connect to a storage device to receive data. For example, when a USB storage device is connected to the input unit, images stored in the USB can be copied to the memory and stored or displayed on the display.

According to one embodiment of the present disclosure, the input unit can receive an external operation. For example, when the jog included in the input unit is manipulated for the control of the camera unit, the camera unit can change the shooting position or change the focus. The input unit may include a switch for changing the status judging photograph. The input unit may receive inputs of various users.

In accordance with one embodiment of the present disclosure, an input may include or be coupled to one or more external input devices. The user can connect the external storage medium to the input unit and perform firmware upgrade, content storage, and the like of the device 100. [ In addition, a user may connect a keyboard, a mouse, and the like to the input unit. In addition, the USB port of the input unit can support USB 2.0, and USB 3.0. The input unit may include an interface such as a thumb bolt as well as a USB port. In addition, the input unit may receive external inputs by combining various input devices such as a touch screen, a button, etc., and the device 100 may be controlled by the control unit in response to the input.

According to embodiments of the present disclosure, the apparatus 100 may further include a display. The display can display an image photographed by the photographing module 111. The display may display the image currently being photographed by the photographing module 111 or may display an image stored in the memory unit. For example, the display according to the embodiments of the present disclosure may display image data that is matched by the image data matching unit 150, or restored by the data restoring unit 150. This allows the user to monitor the displayed image.

Further, the display can output a result of state determination for the target. Such a display may receive and output an image of an external device, and may output an image state determination result of the external device.

According to embodiments of the present disclosure, the display may output a photographable photographing mode. For example, you can output 200dip, 300dip, and 600dip image resolution. In addition, a selection screen can be displayed so that an image size, an image shooting mode, a photographing apparatus, and the like can be selected. The display may comprise any visual output device, such as an LCD, LED, CRT, or the like. The display examples and components described above are merely illustrative, and the present disclosure is not limited thereto. Also, the display may be provided as a touch screen to receive input through a user's touch, and may operate as a user input interface.

Apparatus 100 in accordance with embodiments of the present disclosure may be configured to enable acoustic output. It is also possible to output a warning sound in accordance with the result of the state determination or not.

A controller in accordance with embodiments of the present disclosure can communicate with all of the other components described above, and can organically control their operation.

The various embodiments described herein may be embodied in a recording medium readable by a computer or similar device using, for example, software, hardware, or a combination thereof. According to a hardware implementation, the embodiments described herein may be implemented as application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays May be implemented using at least one of a processor, controllers, micro-controllers, microprocessors, and other electronic units for performing other functions. In some cases, The embodiments described may be implemented by the control itself.

According to a software implementation, embodiments such as the procedures and functions described herein may be implemented with separate software modules. Each of the software modules may perform one or more of the functions and operations described herein. Software code can be implemented in a software application written in a suitable programming language. The software code is stored in the memory 190 and can be executed by the control unit.

12 is a conceptual diagram illustrating an example of a neural network according to embodiments of the present disclosure.

Throughout this specification, a neural network, a network function, and a neural network may be used interchangeably. A neural network may consist of a set of interconnected computational units, which may be referred to generally as " nodes ". These " nodes " may also be referred to as " neurons ". The neural network comprises at least two nodes. The nodes (or neurons) that make up the neural network may be interconnected by one or more " links ".

Within a neural network, two or more nodes connected through a link may form a relationship of input and output nodes relatively. The concepts of the input node and the output node are relative, and any node in the output node relationship with respect to one node can be in the input node relationship with the other node, and vice versa. As described above, the input node-to-output node relationship can be generated around the link. One or more output nodes may be connected to one input node via a link, and vice versa.

In an input node and an output node relationship connected through one link, the output node can be determined based on data input to the input node. Here, a node interconnecting an input node and an output node may have a weight. The weights may be variable and may be varied by the user or algorithm to perform the desired function of the neural network. For example, if one or more input nodes to one output node are interconnected by respective links, then the output node is set to the values input to the input nodes associated with the output node and to the corresponding links to the respective input nodes The output node value can be determined based on the weight.

As described above, the neural network is interconnected by two or more nodes over one or more links to form an input node and an output node relationship within the neural network. The characteristics of the neural network can be determined according to the number of nodes and links in the neural network, the association between the nodes and links, and the value of the weight assigned to each of the links. For example, if there are the same number of nodes and links, and there are two neural network networks with different weight values between the links, then the two neural network networks may be recognized as being different from each other.

As shown in FIG. 12, the neural network may be configured to include two or more nodes. Some of the nodes that make up the neural network may configure one layer based on distances from the original input node. For example, a set of nodes with a distance n from the first input node may be , n layers can be constructed. The distance from the original input node may be defined by the minimum number of links that must go through to reach that node from the original input node. However, the definition of this layer is arbitrary for explanation, and the degree of the layer in the neural network can be defined in a manner different from the above. For example, the layer of nodes may be defined by distance from the final output node.

The first input node may refer to one or more nodes in the neural network network to which data is directly input without going through a link in relation to other nodes. Or may refer to nodes in a neural network that do not have other input nodes connected by a link in the relationship between the nodes based on the link. Similarly, the final output node may refer to one or more nodes that do not have an output node, in relation to other ones of the nodes in the neural network. Also, the hidden node may refer to nodes constituting a neural network other than the first input node and the last output node.

In the exemplary embodiment shown in FIG. 12, the neural network may have three layers. The three layers may include an initial input layer 210, a hidden layer 220, and a final output layer 230, in accordance with the above definition. However, the exemplary neural network shown in Fig. 12 is simplified for explanation, and the number of layers and the number of nodes constituting the neural network can be variously set according to the purpose and complexity of the neural network. In other words, there may be a neural network that includes two or more hidden layers (not shown). In addition, the number of nodes included in each layer may be more or less than that shown in FIG.

At the initial input layer 210, information about an object to be determined by the neural network may be input to each of the nodes 211, 212, and 213 constituting the initial input layer 210. The number of input nodes of the initial input layer 210 may be different depending on the type of the neural network. In one example, if the object to be determined by the neural network is image data, the input data extracted from the image data may be input to the initial input layer 210 of the neural network. In a different neural network having the same purpose of determining image data, the number of input nodes (not shown) of the neural network may be determined differently depending on the purpose, design, accuracy, etc. of the neural network.

For example, a single neural network may include a corresponding number of input nodes (not shown) in the original input layer 210 of the number of optical information for each of the pixels making up the image data. For example, in the case of a neural network for determining square image data having a size of 128 x 128 in the horizontal and vertical directions, information about the R, G, B numerical values for each pixel is stored in the first input layer 210). Thus, the original input layer may include 128 X 128 X 3 input nodes. In another example, a neural network network for determining image data may include fewer nodes in the original input layer 210 than the example described above. For example, numerals (e.g., the number of consecutive pixels on the image data, the number of inflection points, etc.) indicative of the characteristics of the image data to be determined may be input to the initial input layer.

The neural network networks used in the embodiments disclosed herein are not limited to neural network networks for determining image data, and neural network networks may include various data (e.g., nodes on a computer network) Data included in a computer file to be compressed, data indicating a physical state of a mechanical component in robotics engineering, financial analysis data, etc.) can be processed as input data.

In the neural network 200, data input to one n layer can be output as n + 1 layers. Nodes included in the n-th layer can be linked with nodes included in the n + 1-th layer. One node included in the n + 1 layer may include as input values a value computed based on weights assigned to one or more nodes and one or more links included in an n-layer connected via a link with the node have. In other words, k nodes (Nnode ₁ , Nnode ₂ , Nnode ₃ , ... Nnode _k ) included in the n layer and one node (N + 1) node ₁ included in the n + (L ₁ , ..., L _k ), the following equation (5) can be established.

Here, Value (X) may mean a node value input to a node X existing in a neural network. Weight (L) may refer to the weight of link L existing in the neural network. y = f (x) may mean that the y value has a functional relationship determined based on the x factor.

Other nodes included in the (n + 1) -th layer may also have input values as values calculated based on the values of the nodes included in the n-th layer and the weights of the links. In other words, the node values to be input to the nodes included in the (n + 1) -th layer are the values of the nodes included in the n-th layer and the connection relations of the links connecting the nodes of the n- And their weights.

As described above, the initial input layer 210 may be defined as a zero layer. Therefore, the initial input layer 210 and the hidden layer 220 of FIG. 12 are the 0-layer and the 1-layer, and are included in the 0-layer (the initial input layer 210) according to the calculation method inside the above- The value of the nodes 221, 222, 223, and 224 included in one layer (the hidden layer 220) is calculated based on the weight of the data and links 215 input to the nodes 211, 212, Can be determined. Similarly, the values of the nodes 231 and 232 included in the two layers (the final output layer 230) are the same as those of the nodes 221, 222, 223, and 224 included in the first layer 220, 225). ≪ / RTI >

It is apparent to those skilled in the art that the neural network 200 shown in FIG. 12 is only an example of a neural network, and that various types of neural networks exist. A neural network may be referred to as a network function in the sense that input and output values are present. In other words, in this specification, the network function may be understood to mean a neural network.

In this specification, a network function, a neural network, can be used interchangeably. Also, the network function and the neural network may be understood to include an auto encoder, a diabolo network, and the like. When two or more neural network networks exist, if more than one neural network output nodes are input nodes of another neural network, two or more neural network networks may be referred to as a serial connection. When two or more neural network networks are connected in series, two or more serially connected neural network networks can constitute a neural network network.

On the other hand, objects of the inspection target in the process are imaged and inspected through the neural network. For example, an imaged textile fabric may be inspected during fabrication to find scratches, and various objects may be imaged and inspected to find defects in the material's holes or semiconductor circuits. There is a limitation in detecting and analyzing image data by a person, and a method for determining the state of image data with high accuracy is required in the art.

Those of ordinary skill in the art will understand that information and signals may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips that may be referenced in the above description may include voltages, currents, electromagnetic waves, magnetic fields or particles, Particles or particles, or any combination thereof.

Those skilled in the art will appreciate that the various illustrative logical blocks, modules, processors, means, circuits, and algorithm steps described in connection with the embodiments disclosed herein may be implemented or performed with a specific purpose, (Which may be referred to herein as "software") or a combination of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the design constraints imposed on the particular application and the overall system. Those skilled in the art may implement the described functions in various ways for each particular function, but such implementation decisions should not be interpreted as being outside the scope of the present disclosure.

The various embodiments presented herein may be implemented as a method, apparatus, or article of manufacture using standard programming and / or engineering techniques. The term "article of manufacture" includes a computer program, carrier, or media accessible from any computer-readable device. The medium may comprise a storage medium or a communication medium. For example, the computer-readable storage medium can be a magnetic storage device (e.g., a hard disk, a floppy disk, a magnetic strip, etc.), an optical disk (e.g., CD, DVD, But are not limited to, memory devices (e. G., EEPROM, card, stick, key drive, etc.).

Communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal, such as a carrier wave or other transport mechanism, Media. The term modulated data signal refers to a signal that has one or more of its characteristics set or changed to encode information in the signal. By way of example, and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, or other wireless media.

It will be appreciated that the particular order or hierarchy of steps in the presented processes is an example of exemplary approaches. It will be appreciated that, based on design priorities, a particular order or hierarchy of steps in the processes may be rearranged within the scope of this disclosure. The appended method claims provide elements of the various steps in a sample order, but are not meant to be limited to the specific order or hierarchy presented.

The description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present disclosure should not be construed as limited to the embodiments set forth herein, but is to be accorded the widest scope consistent with the principles and novel features presented herein.

Claims (20)

A method of correcting target image data,
Generating reference image data based on a plurality of normal image data without defects;
Detecting one or more reference image feature points from the reference image data;
Detecting one or more target image feature points from the target image data;
Determining a target image estimating feature point and a reference image estimating feature point for each of the reference image data and non-detected feature points and unmatched feature points from the target image, based at least in part on an algorithm for feature point estimation;
Matching the reference image data and the target image data based at least on the reference image feature point, the target image feature point, and the estimated feature point; And
Restoring at least a portion of the data of the target image data lost by the illumination;
/ RTI >
A method for correcting target image data.
The method according to claim 1,
Wherein the detecting of the one or more reference image feature points and the one or more target image feature points comprises:
Dividing the reference image data and the target image data into at least one region having a predetermined size, and then detecting a feature point within each region;
/ RTI >
A method for correcting target image data.
3. The method of claim 2,
The predetermined size may be,
Wherein the reference image data and the target image data have a size set based on at least one of image resolution, size, color mode, and aspect ratio of the target image data.
A method for correcting target image data.
The method according to claim 1,
Wherein the step of detecting feature points from the reference image data comprises:
Detecting a feature point candidate group from the reference image data based at least on a first algorithm for feature point detection; And
Detecting a feature point necessary for matching with the target image data from the feature point candidate group based at least on a second algorithm for feature point detection;
≪ / RTI >
A method for correcting target image data.
The method according to claim 1,
Wherein detecting at least one feature point from the target image data comprises:
Detecting a feature point candidate group from the target image data by determining position coordinates on the target image data corresponding to position coordinates for each of the feature points detected from the reference image data as a feature point; And
Detecting a feature point necessary for matching with the reference image data from the feature point candidate group based at least on a second algorithm for feature point detection;
≪ / RTI >
A method for correcting target image data.
5. The method of claim 4,
Wherein the first algorithm for the feature point detection comprises:
Which is at least one of a Harris Corner Detector algorithm, a Shi-Tomasi Corner Detector algorithm, and a SIFT_DoG algorithm,
A method for correcting target image data.
The method according to any one of claims 4 to 5,
Wherein the second algorithm for the feature point detection comprises:
Normalized cross correlation (NCC) result data between the reference image data and the second image data associated with the reference image data and between the target image data and the second image data associated with the target image data, The Super Resoulution Feature Matching algorithm,
A method for correcting target image data.
8. The method of claim 7,
Wherein the second image data comprises:
And generating at least one of distortion and loss for the image data,
A method for correcting target image data.
8. The method of claim 7,
Wherein the second image data comprises:
A Gaussian blur filter and a Gaussian noise applied to the image data,
A method for correcting target image data.
The method according to claim 1,
The algorithm for the minutiae estimation includes:
A plurality of feature points, at least based on the number of feature points and the distance connecting each feature point, and corresponding to the number of feature points,
A method for correcting target image data.
11. The method of claim 10,
The conversion model equation,
If the number of the minutiae is one, offset conversion,
If the number of feature points is two, similarity transformation,
If the number of feature points is 3, the affine transformation and
If the number of feature points is 4,
A method for correcting target image data.
The method according to claim 1,
Wherein matching the reference image data and the target image data comprises:
Transforming the target image data; And
Normalizing the size of the target image data;
≪ / RTI >
A method for correcting target image data.
The method according to claim 1,
Determining status information of the target image data using the learned network function for the reference image data;
≪ / RTI >
A method for correcting target image data.
14. The method of claim 13,
Wherein the status information of the target image data includes:
And determining, based on state information for each of the one or more areas divided into a predetermined size,
A method for correcting target image data.
An apparatus for correcting target image data,
A reference image generated based on a plurality of normal image data without defects and an image data acquiring unit acquiring the target image data;
A first detector for detecting one or more feature points from the reference image data;
A second detecting unit detecting at least one feature point from the target image data;
An estimated feature point determiner for determining an estimated feature point for each of the feature point and the unmatched feature point that are not detected from the reference image data and the target image data, based at least on an algorithm for feature point estimation;
An image data matching unit for matching the reference image data and the target image data based at least on the feature points and the estimated feature points; And
A data restoring unit for restoring at least a part of the data of the target image data lost by illumination;
/ RTI >
A device for correcting target image data.
16. The method of claim 15,
Wherein the image data obtaining unit comprises:
A photographing module for photographing a target;
A first acquiring module for acquiring a plurality of normal image data without defects; And
A reference image generation module that generates the reference image data based on the plurality of normal image data;
/ RTI >
A device for correcting target image data.
16. The method of claim 15,
Wherein the image data obtaining unit comprises:
A photographing module for photographing a target; And
A second acquiring module for acquiring the reference image generated based on a plurality of normal image data without defects;
/ RTI >
A device for correcting target image data.
16. The method of claim 15,
Wherein the image data obtaining unit comprises:
A third acquiring module for acquiring the reference image data generated based on the plurality of normal image data without defects and the target image data photographed by the photographing module;
/ RTI >
A device for correcting target image data.
16. The method of claim 15,
A status information determination unit for determining status information of the target image data using the network function learned for the reference image data;
≪ / RTI >
A device for correcting target image data.
A computer program stored in a computer readable storage medium comprising a plurality of instructions executed by one or more processors of an apparatus for correcting target image data,
The computer program comprising:
Obtaining a reference image and a target image generated based on a plurality of normal image data that are free from defects;
Detecting one or more feature points from the reference image data;
Detecting one or more feature points from the target image data;
Determining an estimated feature point for each of the feature point and non-matched feature point that are not detected from the reference image data and the target image data, based at least on an algorithm for feature point estimation;
Instructions for matching the reference image data and the target image data based at least on the feature points and the estimated feature points; And
Restoring at least a portion of the data of the target image data lost by illumination;
/ RTI >
A computer program stored on a computer readable storage medium.

Images