KR101972428B1 - Apparatus and method for processing image with mode selection for protecting privacy - Google Patents

Abstract

The present invention relates to an image processing apparatus capable of setting a mode for personal information protection and a method thereof. More specifically, according to an aspect of the present invention, image processing apparatus, which is capable of setting a mode for personal information protection, includes: an image input unit for receiving images composed of multiple frames captured by a camera; an object recognition unit for recognizing an object area corresponding to a vehicle and an object area corresponding to a person in the images input through the image input unit; a vehicle license plate number recognition unit which extracts characters from the object areas recognized by the object recognition unit by using a comparison with preset templates and recognizes the vehicle license plate number based on the extracted characters; a mode setting unit for setting a mode setting state to be any one of an unspecified mode, a specific mode, and a normal mode in accordance with an input from a user; a list unit for storing information related to multiple vehicle license plate numbers in the form of a list; and a masking unit which converts each pixel value in the object areas recognized by the object recognition unit to be different from existing values by using preset patterns based on the mode setting state of the mode setting unit and whether the vehicle license plate number recognized by the vehicle license plate number recognition unit corresponds to a vehicle license plate number included in the vehicle license plate number information of the list unit.

Description

 TECHNICAL FIELD [0001] The present invention relates to an image processing apparatus and method capable of setting a mode for privacy protection,

The present invention relates to an image processing apparatus and method capable of setting a mode for protecting personal information, and more particularly, to an image processing apparatus and method capable of setting a mode for personal information protection And more particularly, to a video processing apparatus and method capable of setting a video signal.

Video surveillance using CCTV is widely used for urban facility management and policing. In recent years, intelligent control systems based on image processing and object recognition for images input through cameras have been introduced.

In the image collected by the camera, it is possible to recognize objects such as a person or a vehicle through pattern recognition or the like. In collecting information about a person, there are legal restrictions due to protection of personal information and the like.

Various cases are generated for each purpose in performing a video control such as a case where a specific vehicle is picked up, a vehicle which should not be exposed, a case where information about a person should not be included in a recognized object, and so on. There is a problem that the masking process is performed collectively for protecting the privacy of the object information to be noticed in the video control method according to the related art or the personal information other than the object information to be learned is included and the corresponding image can not be used as it is, There was a limitation that it could not cope with different video control purposes.

KR 10-1487461 B1 KR 10-2014-0012474 A KR 10-1381343 B1 KR 10-1863727 B1

It is an object of the present invention to provide a method and a system for setting a mode setting state into one of an unspecified mode, a specific mode, and a normal mode according to a user's input, and to perform masking of object areas according to a mode setting state And an image processing apparatus and method capable of setting a mode for the image processing apparatus.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood from the following description.

According to an aspect of the present invention, there is provided an image processing apparatus capable of setting a mode for privacy protection according to an embodiment of the present invention includes a video input unit for inputting an image composed of a plurality of frames captured by a camera, An object recognizing unit for recognizing an object area corresponding to a vehicle and an object area corresponding to a person in an input image; a character extracting unit for extracting characters from the object area recognized by the object recognizing unit A mode setting unit for setting the mode setting state in one of an unspecified mode, a specific mode, and a normal mode according to a user's input; and a mode setting unit for storing a plurality of vehicle number information in a list form And the vehicle number recognized by the vehicle number recognizing unit is the vehicle number of the list part And a masking unit for transforming each of the pixel values of the object regions recognized by the object recognizing unit by using a preset pattern differently from an existing value, based on whether or not the vehicle number belonging to the information is associated with the mode setting state of the mode setting unit .

According to another aspect of the present invention, there is provided an image processing method capable of setting a mode for protecting personal information, including: receiving an image composed of a plurality of frames captured from a camera; Recognizing a corresponding object area; extracting a character from the recognized object area in comparison with a preset template; recognizing the vehicle number based on the extracted character; determining a mode setting state as an unspecified A first list in which only vehicle numbers requiring disclosure of a plurality of pieces of vehicle number information are classified and a second list in which only vehicle numbers that need to be privately classified are classified; Storing the recognized vehicle number in a list form, and storing the recognized vehicle number in the list of vehicle numbers of the first list or the second list Converting each of the pixel values into a different value from the existing value using the predetermined pattern, based on whether or not the vehicle number to which the vehicle belongs and the mode setting state are recognized.

According to the present invention, in image recognition through a camera, a non-specific mode, which is a mode for displaying only vehicle numbers belonging to a predetermined list and masking the remaining vehicle numbers and objects for information protection, There is an advantage that it is possible to set a specific mode which is a mode for preventing exposure and a normal mode which is a mode for masking the object area to enable information protection.

According to the present invention, in a case where the amount of calculation is large due to recognition and masking processing of the car number information, or in the case where hardware for calculation processing is not supported, a plurality of car number information Can be performed.

Further, according to the present invention, it is possible to use a priority code having a binarized value, a list binarization value, a binarization value of car number information, and use a characteristic map according to a convolutional neural network, Can be achieved.

1 is a block diagram illustrating an image processing apparatus capable of setting a mode for privacy protection according to an embodiment of the present invention.
2 is a flowchart illustrating an image processing method capable of setting a mode for personal information protection according to another embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described below, but may be embodied in various different forms, and these embodiments are not intended to be exhaustive or to limit the scope of the present invention to the precise form disclosed, It is provided to inform the person completely of the scope of the invention. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Hereinafter, an image processing apparatus and method capable of setting a mode for privacy protection according to embodiments of the present invention will be described with reference to FIGS. 1 and 2. FIG. FIG. 1 is a block diagram illustrating an image processing apparatus capable of setting a mode for privacy protection according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating an image processing apparatus according to another embodiment of the present invention, Fig.

An image processing apparatus 10 capable of setting a mode for privacy protection according to an embodiment of the present invention includes an image input unit 100, an object recognition unit 200, a vehicle number recognition unit 300, A display section 700, a highlight section 800, a vehicle characteristic map section 910, and a human characteristic map section 920. The display section 700 includes a display section 700, a list section 500, a masking section 600, a display section 700, .

The image input unit 100 receives an image composed of a plurality of frames captured by the camera (S100).

The object recognition unit 200 recognizes an object region corresponding to a vehicle and an object region corresponding to a person in an image input through the image input unit 100 (S200).

The vehicle number recognizing unit 300 extracts a character from the object area recognized by the object recognizing unit in comparison with a preset template, and recognizes the vehicle number based on the extracted character (S300).

The mode setting unit 400 divides the mode setting state into one of an unspecified mode, a specific mode, and a normal mode according to a user's input (S400).

In the image processing apparatus 10 capable of setting a mode for protecting personal information according to an embodiment of the present invention, the unspecified mode displays only the vehicle numbers belonging to the predetermined list, and the remaining vehicle numbers and objects are used for information protection And the specific mode is a mode for masking the vehicle numbers belonging to the predetermined list, which is a mode for preventing the number of the VIP vehicle from being exposed, and the normal mode is a mode for masking all the object areas So that information can be protected.

Here, the mode setting unit 400 may be configured to distinguish the mode setting state from the unspecified mode, the specific mode, and the normal mode according to at least two binary values.

For example, in the mode setting unit 400, the unspecified mode may be set to 01, the specific mode may be set to 10, and the normal mode may be set to a value of 11.

The list unit 500 stores a plurality of pieces of car number information in the form of a list, and includes a first list in which only car numbers requiring disclosure of a plurality of pieces of car number information are classified, and a second list in which only car numbers requiring privatization are classified (S500).

Here, the list unit 500 may divide the first list and the second list according to the binarized values, converts a plurality of car number information into binarized values, stores each car number information in a first list And to store the binarized values so as to correspond to any of the first list and the second list. In this case, in the list unit 500, the car number information may be converted into a binarization decimal notation only in the numeric portion and stored. In this case, since 4 bits are required for each number in order to store the car number information, ) May be configured to allocate at least one bit for distinguishing the first list from the second list and at least 16 bits for storing car number information according to the binarized value to generate list data.

In addition, the list unit 500 may have a code for assigning priorities when a plurality of car number information are stored, and generating list data by allocating bits for writing a code for prioritization to upper bits .

Here, the code for prioritization may be stored in accordance with the user's input or stored in advance.

Meanwhile, the car number information may be converted and stored in the form of a BCD code.

Table 1 below shows an example in which 2 bits are stored to store the binarization value according to the list classification, and 24 bits are allocated to store the binarization value of the car number information. Only the least significant bit may be used for distinguishing the first list and the second list. Therefore, it is necessary to allocate at least one bit. In the case of the car number information, You will need to allocate 16 bits.

number List division the car's number Priority code List binarization value Binarization value of car number information One First List 70 you 5221 001 01 011100000101001000100001 2 First List 40 is 4231 100 01 010100000100001000110001 3 Second list 30 is 3123 110 10 001100000011000100100011 4 Second list 52 is 1151 001 10 010100100001000101010001 5 First List 24, 5531 001 01 001001000101010100110001 6 Second list 60, 6631 001 10 011000000110011000110001

In the case of Table 1, the list unit 500 includes the binarization value of the car number information, the priority bin code, the list binarization value, and the binarization value of the car number information for the car number ' And the value of '00101011100000101001000100001' may be recorded in the content part of the list data, not in the header part. When the car number '30' belonging to the second list is' 3123 ',' 11010001100000011000100100011 'is recorded as the binarization value. If the binarization value assigned to the upper bit is' If the car number '30 is 3123', it is recorded as' 110 '. Therefore, the car number' 30 'may be calculated or masked first because the car number' 3123 'takes precedence over the car number'

In an image processing apparatus and method capable of setting a mode for protecting personal information according to embodiments of the present invention, when there is a large amount of computation due to recognition and masking processing of the car number information, or hardware for arithmetic processing is not supported The calculation for the plurality of car number information is performed according to the code for prioritization so that important car number information can be processed first.

The masking unit 600 determines whether the vehicle number recognized by the vehicle number recognizing unit 300 corresponds to the vehicle number belonging to the vehicle number information of the list unit 500 and the mode setting state of the mode setting unit 400 In operation S600, object regions recognized by the object recognition unit 200 are converted into pixel values different from the existing values using a predetermined pattern.

The masking unit 600 determines whether to convert the pixel values of the object regions using the mode setting state classified by the binarized values of the mode setting unit 400 and the list data of the binarized values generated in the list unit 500 It can be decided.

The masking unit 600 receives the list data of the list unit 500, the input image of the image input unit 100, and the data of the object area recognized by the object recognition unit 200, And a program logic or multiplexer for outputting an output image of a size corresponding to the input image.

If the mode setting state is the non-specific mode and the vehicle number recognized through the vehicle number recognizing unit 300 corresponds to the vehicle number belonging to the vehicle number of the list unit 500, the masking unit 600 The object region may be configured to maintain the existing pixel values and to convert each of the pixel values of the remaining object regions recognized by the object recognition unit 200 to different values from the existing values.

If the mode setting state is the specific mode and the vehicle number recognized by the vehicle number recognizing unit 300 corresponds to the vehicle number belonging to the vehicle number of the list unit 500, It may be possible to convert each pixel value of the object region differently from the existing value.

In addition, the masking unit 600 may convert pixel values of object regions recognized by the object recognition unit 200 differently from the existing values, if the mode setting state is the normal mode.

On the other hand, if the vehicle number recognized by the vehicle number recognizing unit 300 corresponds to the vehicle number belonging to the second list, the masking unit 600 may determine whether or not the pixel number of the object region corresponding to the vehicle number It may be possible to convert each of the values differently from the existing value.

The display unit 700 displays the masked image of the object regions recognized in the image input through the image input unit 100 according to the mode setting state of the mode setting unit (S700).

The display unit 700 includes an image input unit 100, an object recognition unit 200, and a masking unit 100. The image input unit 100 includes an image input unit 100, an image input unit 100, 600 and may be provided with masked pixel values for the object regions.

The highlight unit 800 generates a highlight mark for the object area corresponding to the vehicle number when the vehicle number recognized through the vehicle number recognizing unit corresponds to the vehicle number belonging to the first list. On the other hand, the display unit 700 may overlap the highlight mark for the object area corresponding to the car number in the masked image (S800).

According to the embodiments of the present invention, highlight mark generation is performed in the highlight unit 800, so that even when the mode setting state is the specific mode or the normal mode, an object corresponding to the vehicle number belonging to the first list in the display unit 700 There is an advantage of knowing the area.

The vehicle characteristic map unit 910 stores and manages a vehicle image feature map obtained by performing a convolution operation using a plurality of pattern masks preset for a plurality of vehicle images, ).

Meanwhile, the vehicle characteristic map unit 910 may store and manage a vehicle image feature map obtained by learning through a convolution neural network using a plurality of pattern masks preset for a plurality of vehicle images.

The human feature map unit 920 stores and manages a human image feature map obtained by performing a convolution operation using a plurality of pattern masks preset for a plurality of human images, ).

The human feature map section 920 may store and manage the human image feature map obtained by learning through the convolutional neural network using a plurality of pattern masks preset for a plurality of human images.

The vehicle characteristic map portion 910 and the human characteristic map portion 920 may have characteristic maps each having a high recognition rate of the vehicle and the human being through the convolutional neural network.

On the other hand, the object recognition unit 200 may calculate a convolution operation result of the image input through the image input unit 100 using a plurality of preset pattern masks as a predetermined vehicle image feature map or a predetermined person image feature map It is possible to recognize an area having similarity in an image image as an object area if it is determined that the similarity is within a preset range.

According to embodiments of the present invention, it is possible to use a priority code having a binarized value, a list binarization value, a binarization value of car number information, and use a characteristic map according to a convolutional neural network, So that an operation is performed. Therefore, when used together with the FPGA-based image processing hardware, the processing speed can be improved, and the input / output data can be binarized and utilized, which makes it possible to use more intuitively.

It will be understood by those skilled in the art that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of the present invention is defined by the appended claims rather than the detailed description, and all changes or modifications derived from the scope of the claims and their equivalents should be construed as being included in the scope of the present invention.

10: image processing device capable of mode setting for privacy protection,
100: image input unit, 200: object recognition unit, 300: car number recognition unit,
400: mode setting unit, 500: list unit, 600: masking unit, 700:
800: highlight section, 910: vehicle characteristic map section, 920: human characteristic map section.

Claims (13)

A video input unit that receives an image composed of a plurality of frames captured from a camera;
An object recognition unit for recognizing an object region corresponding to a vehicle and an object region corresponding to a person in an image input through the image input unit;
A car number recognition unit for extracting characters from the object area recognized by the object recognition unit in comparison with a predetermined template and recognizing the car number based on the extracted character;
A mode setting unit configured to set the mode setting state to one of an unspecified mode, a specific mode, and a normal mode according to a user's input;
A list unit for storing a plurality of car number information in a list form, the list including a first list in which only car numbers requiring disclosure of a plurality of car number information are classified, and a second list in which only car numbers requiring privately are sorted;
Based on whether the vehicle number recognized by the vehicle number recognizing unit corresponds to the vehicle number belonging to the vehicle number information of the list unit and the mode setting state of the mode setting unit, A masking unit for converting each of the pixel values to an existing value; And
A highlighting unit for generating a highlight mark for an object area corresponding to the vehicle number when the vehicle number recognized through the vehicle number recognizing unit corresponds to the vehicle number belonging to the first list;
And a mode setting unit for setting a mode for privacy protection. The image forming apparatus according to claim 1,
If the mode setting state is the unspecified mode and the vehicle number recognized through the vehicle number recognizing unit corresponds to the vehicle number belonging to the vehicle number of the list unit, the object region corresponding to the vehicle number holds the existing pixel values, The pixel values of the remaining object regions recognized by the object recognition unit are converted differently from the existing values,
When the mode setting state is the specific mode and the vehicle number recognized by the vehicle number recognizing unit corresponds to the vehicle number belonging to the vehicle number of the list portion, the pixel value of the object region corresponding to the vehicle number is divided into an existing value To convert differently,
If the mode setting state is the normal mode, converting the pixel values of the object regions recognized by the object recognition unit to different values from the existing values
Image processing apparatus capable of setting a mode for protecting personal information. delete delete The image forming apparatus according to claim 1,
When the vehicle number recognized by the vehicle number recognizing unit corresponds to the vehicle number belonging to the second list, the pixel values of the object region corresponding to the vehicle number are different from the existing values regardless of the mode setting state Conversion
Image processing apparatus capable of setting a mode for protecting personal information. The apparatus of claim 1,
A result of a convolution operation using a plurality of preset pattern masks for an image input through the image input unit is compared with a predetermined vehicle image feature map or a preset human image feature map to have similarity within a predetermined range And recognizing the region having the degree of similarity as the object region in the image if it is determined
Image processing apparatus capable of setting a mode for protecting personal information. The method according to claim 1,
A display unit for displaying a masked image of object regions recognized in an image input through the image input unit according to a mode setting state of the mode setting unit;
And a mode setting unit for setting a mode for personal information protection. The display device according to claim 7,
And displaying the highlighted mark for the object area corresponding to the car number in the masked image in an overlapping manner
Image processing apparatus capable of setting a mode for protecting personal information. The method according to claim 1,
A vehicle feature map portion having a vehicle image feature map obtained by convolution operation using a plurality of pattern masks preset for a plurality of vehicle images; And
A human feature map section having a human image feature map obtained by convolution operation using a plurality of pattern masks preset for a plurality of human images;
Further comprising:
The object recognition section recognizes the object area based on the vehicle image feature map of the vehicle feature map section or the human image feature map of the human feature map section
Image processing apparatus capable of setting a mode for protecting personal information. Receiving an image composed of a plurality of frames captured from a camera;
Recognizing an object region corresponding to a vehicle and an object region corresponding to a person in an input image;
Extracting a character from the recognized object area in comparison with a preset template, and recognizing the vehicle number based on the extracted character;
Setting the mode setting state into one of an unspecified mode, a specific mode, and a normal mode according to a user's input;
Storing a first list in which only a plurality of vehicle number information is classified, the second list being divided into a first list and a second list;
Whether or not the recognized vehicle number corresponds to the vehicle number belonging to the vehicle number information of the first list or the second list and the mode setting state, To a value different from an existing value; And
Displaying a masked image of object regions recognized in an input image according to the mode setting state; / RTI >
Wherein the displaying comprises:
Overlapping and displaying a highlight mark for an object area corresponding to a vehicle number belonging to the first list in a masked image;
And a mode setting unit configured to set a mode for personal information protection. 11. The method of claim 10,
When the mode setting state is the unspecified mode and the recognized vehicle number corresponds to the vehicle number belonging to the first list, the object region corresponding to the vehicle number maintains the existing pixel values and the pixel values Each of which is different from the existing value,
Wherein when the mode setting state is a specific mode and the recognized vehicle number corresponds to a vehicle number belonging to the second list, each pixel value of an object region corresponding to the vehicle number is converted differently from an existing value,
If the mode setting state is the normal mode, each of the pixel values of the recognized object regions is converted differently from the existing value
A method of image processing capable of setting a mode for privacy protection. delete delete

Images