Method and apparatus for recognizing ID card

1. A method for recognizing an identification (ID) card of a user terminal using deep learning, the method comprising:

extracting an outline region of the ID card included in an input image using a first neural network model;

modifying an ID card image of the image to a reference form using at least a partial value of the extracted outline region; and

determining whether the modified ID card image is valid according to the presence of a recognition inhibitory factor in the modified ID card image using a third neural network model and recognizing text information in the valid ID card image,

wherein the extracting the outline region of the ID card comprises re-extracting the outline region of the ID card in a re-input image when the recognition inhibitory factor is present.

2. The method of claim 1 , further comprising:

determining whether the ID card exists in the input image using a second neural network model,

wherein the extracting the outline region of the ID card comprises extracting the outline region when the ID card exists.

3. The method of claim 2 , further comprising:

enhancing an outer pixel value of the text information in the modified ID card image against periphery information,

wherein the recognizing the text information comprises recognizing the text information in the ID card image using the enhanced pixel value.

4. The method of claim 1 , wherein,

wherein the recognizing the text information comprises recognizing the text information in the ID card image using a fourth neural network model.

5. The method of claim 1 , wherein

the first neural network model configured to label at least one corner position of the ID card image, and perform learning through data generated by converting the labeled ID card image according to a predetermined condition.

6. The method of claim 1 , wherein

the third neural network model configured to perform learning by collecting the ID card image including the recognition inhibitory factor as learning data.

7. An apparatus for recognizing an identification (ID) card based on deep learning, the apparatus comprising:

at least one processor; and

a non-transitory computer-readable medium storing instructions that when executed by the at least one processor causes the at least one processor to perform operations including:

extracting an outline region of the ID card included in an input image using a first neural network model;

modifying an ID card image of the image to a reference form using at least a partial value of the extracted outline region;

determining whether the modified ID card image is valid according to the presence of a recognition inhibitory factor in the modified ID card image using a third neural network model; and

recognizing text information in the valid ID card image,

wherein the extracting the outline region of the ID card comprises re-extracting the outline region of the ID card in a re-input image when the inhibitory factor determining unit determines the modified ID card is invalid according to the presence of the recognition inhibitory factor.

8. The apparatus of claim 7 , further comprising:

determining whether the ID card exists in the input image using a second neural network model,

wherein the extracting the outline region of the ID card comprises extracting the outline region when the ID card exists.

9. The apparatus of claim 8 , further comprising:

enhancing an outer pixel value of the text information in the modified ID card image against periphery information,

wherein the determining whether the ID card exists in the input image using the second neural network model comprises recognizing the text information in the ID card image using the enhanced pixel value.

10. The apparatus of claim 7 ,

wherein the recognizing the text information comprises recognizing the text information in the ID card image using a fourth neural network model.

11. A method of providing an identification (ID) card recognition service based on deep learning performed in a service providing server, the method comprising:

releasing an application including an ID card recognition model performing at least one convolution operation so that a user terminal uses the application for an image captured by the user terminal;

receiving a result of checking authenticity of an ID card through a certification agency server from the user terminal according to an ID card recognition result of the released ID card recognition model; and

providing a user request service according to the result of checking authenticity of the ID card,

wherein the ID card recognition model comprising:

extracting an outline region of the ID card included in an input image using a first neural network model;

modifying an ID card image of the image to a reference form using at least a partial value of the extracted outline region; and

determining whether the modified ID card image is valid according to the presence of a recognition inhibitory factor in the modified ID card image using a third neural network model,

wherein the extracting the outline region of the ID card comprises re-extracting the outline region of the ID card in a re-input image captured by the user terminal when the recognition inhibitory factor is present.

12. The method of claim 11 , wherein

the ID card recognition model configured to determine whether the ID card exists in the input image using a second neural network model and extract the outline region using the first neural network model when the ID card exists.

13. The method of claim 11 , wherein

the ID card recognition model configured to enhance an outer pixel value of the text information in the modified ID card image against periphery information and recognize text information in the ID card image using the enhanced pixel value.

14. The method of claim 11 , wherein

the ID card recognition model configured to recognize text information in the ID card image using a fourth neural network model.