KR101675796B1 - LookupTable generating method for Image resolution compensation, Image resolution compensating apparatus and method using LUT - Google Patents

Abstract

The present invention discloses a method of generating a lookup table for correcting an image resolution, and an apparatus and method for correcting an image resolution using the method.
An image resolution correcting apparatus of the present invention includes: a photographing unit having a transparent dome cover and photographing a subject in a direction determined by a current pan / tilt zoom value; an image processor for dividing the photographed image into blocks of a predetermined size, A discrete cosine transform unit for discrete cosine transforming the image data, and a resolution correcting unit for increasing a coefficient corresponding to a component having a largest value among horizontal components, vertical components, and diagonal components representing the spatial characteristics of the image in the discrete cosine transform coefficients do.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a lookup table generating method, a method of compensating image resolution,

The present invention relates to a lookup table generation method for image resolution correction, and an apparatus and method for correcting an image resolution using the lookup table.

2. Description of the Related Art Conventional imaging apparatuses, particularly surveillance cameras for CCTV (Closed Circuit TeleVision), are used in a wide range of fields due to increased demand for security and order maintenance. In general, surveillance cameras installed for the purpose of monitoring the passengers and their dynamics can be classified into a dome type, a set type, a bullet type, and a miniature type according to the installation type. Among them, the domed camera adopts the hemispherical shape which is familiar to the interior illumination device as the basic design element, which reduces the feeling of discomfort felt by the observer and has the advantage of being well fitted with the interior of the interior.

However, since the dome cover of the dome type camera has a unique refractive index and transmittance, there is a problem that the resolution of the optical image is lowered.

The present invention provides a method and apparatus capable of correcting the resolution of an image degraded by a dome cover.

According to another aspect of the present invention, there is provided an image resolution correction apparatus comprising: a photographing unit having a transparent dome cover and photographing a subject in a direction determined by a current pan / tilt zoom value; A discrete cosine transform unit that divides the photographed image into blocks of a predetermined size and performs discrete cosine transform of image data of each block; And a resolution corrector for increasing a coefficient corresponding to a component having a largest value among a horizontal component, a vertical component, and a diagonal component indicating spatial characteristics of the image in the discrete cosine transform coefficient.

Wherein the image resolution correcting device corrects a resolution lowering rate indicating a degree of reduction of a discrete cosine transformation coefficient corresponding to a component having the largest value among the horizontal component, the vertical component, and the diagonal component under the condition that the dome cover is absent, And a look-up table that stores the matching result with the pan tilt zoom value.

The resolution correcting unit may multiply the discrete cosine transform coefficient by the inverse of the resolution lowering rate to increase a coefficient corresponding to a component having the largest value among the horizontal component, the vertical component, and the diagonal component.

According to a preferred embodiment of the present invention, there is provided a method of correcting image resolution, comprising: receiving an image obtained by photographing a subject in a direction determined by a current tilt zoom value of a camera having a transparent dome cover; Dividing the input image into blocks of a predetermined size, and performing discrete cosine transform of the image data of each block; And increasing a coefficient corresponding to a component having a largest value among a horizontal component, a vertical component, and a diagonal component indicating spatial characteristics of the image in the discrete cosine transform coefficient.

According to another aspect of the present invention, there is provided a method of generating a lookup table for image resolution correction, the method comprising: capturing a resolution chart by adjusting a pan tilt zoom value of a camera under a condition that a dome cover is absent and a dome cover; Dividing the photographed image into blocks of a predetermined size, and performing discrete cosine transform of image data of each block; Calculating a horizontal component, a vertical component, and a diagonal component from the discrete cosine transform coefficients and arranging the components from the component having the largest value; And calculating a resolution decrease rate indicating a degree of decrease in a discrete cosine transformation coefficient corresponding to a component having the largest value among the horizontal component, the vertical component, and the diagonal component under the condition that the dome cover is not present, .

The present invention can obtain a high resolution image by correcting the resolution of the photographed image by using the resolution lowering rate calculated corresponding to each region of the dome cover as a parameter.

1 is a block diagram schematically showing an image resolution correction apparatus according to a preferred embodiment of the present invention.
FIGS. 2 to 5 are diagrams illustrating a method of generating a lookup table for image resolution correction according to an embodiment of the present invention.
6 is a flowchart illustrating a method of generating a lookup table for image resolution correction according to an exemplary embodiment of the present invention.
FIG. 7 is a flowchart illustrating a method of correcting the resolution of a surveillance image according to an embodiment of the present invention.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically showing an image resolution correction apparatus according to a preferred embodiment of the present invention.

Referring to FIG. 1, an image resolution correction apparatus 10 includes a photographing unit 110, a driving unit 120, and a control unit 130.

The photographing unit 110 is a camera which is fixed on one side of the base member so as to be capable of 360 degree forward panning and 180 degree tilting, and is covered with a hemispherical transparent dome cover. The photographing unit 110 photographs a subject in a direction determined by a set PTZ value.

The driving unit 120 drives the photographing unit 110 by an operation signal and sets the photographing direction and focus of the photographing unit 110 by the PTZ control signal.

The control unit 130 outputs an operation signal and a PTZ control signal for controlling the operation of the driving unit 120 and performs processing such as correction, compression, and the like on the image data input through the photographing unit 110. The control unit 130 includes an input unit 140, a DCT unit 150, a resolution correction unit 160, a quantization unit 170, an encoding unit 180, and a lookup table (LUT) 190.

The input unit 140 processes the analog image signal input from the photographing unit 110, removes the noise, adjusts the amplitude, and converts the digital image data.

The DCT unit 150 divides the image data into blocks each having a predetermined size, for example, an 8x8 block (hereinafter, referred to as a DCT block), and performs discrete cosine transform (Discrete Cosine Transform) process. Through the discrete cosine transform process, the image data of 8x8 DCT block is transformed into 8x8 DCT coefficients.

The resolution correction unit 160 selectively increases a DCT coefficient corresponding to a component having a largest value among a horizontal component, a vertical component, and a diagonal component indicating spatial characteristics of an image.

The dome cover is a transparent hemisphere that surrounds the PTZ camera and has a unique refractive index and transmittance, and functions as an LPF (Low Pass Filter). Accordingly, the image obtained by photographing the subject with the PTZ value set by the photographing unit 110 is an image formed by light passing through a specific region of the dome cover, and the loss of image information in the high frequency region due to the refractive index and the transmittance of the dome cover And the image resolution may be degraded.

In compensating the degradation of the image caused by the dome cover in the surveillance PTZ camera, the resolution enhancement due to the uniform DCT coefficient increase in the high frequency region of the image is not reflected in the spatial characteristics of the image, . The present invention can improve the resolution by detecting in advance how the dome cover degrades the spatial characteristics of the image by learning and selectively increasing any DCT coefficients from the dome cover in order to find out whether the substantial resolution improvement is achieved.

Specifically, the resolution correction unit 160 multiplies the DCT coefficient by the inverse of the resolution lowering rate T to increase the DCT coefficient in the high frequency region. The resolution lowering rate (T) is a value indicating the resolution degradation of the image due to the dome cover. That is, the resolution decrease rate T indicates how much the DCT coefficient corresponding to the component having the largest value among the horizontal component, the vertical component, and the diagonal component decreases under the condition that the dome cover is present under the condition without the dome cover. Therefore, when the DCT coefficient calculated for the current DCT block of the input image is multiplied by the inverse of the resolution decreasing rate T, the DCT coefficients of all the high frequency regions of the current DCT block are not corrected, but only the specific DCT coefficients are selectively corrected , It is possible to perform resolution correction considering the spatial characteristics of the image.

The quantization unit 170 performs a process of quantizing the corrected DCT coefficients.

The encoding unit 180 compresses the quantized coefficients using a run length coding (RLC) or a variable length coding (VLC). The data compressed by the DCT block is buffered and then transmitted to the outside.

The lookup table (LUT) 190 is information that matches the PTZ value of the dome camera with the resolution decrease rate T, and is stored in the memory. Since the dome cover is not made entirely of the same thickness and the uniformity is also lowered, the resolution degradation depends on the position (PTZ value) of the dome camera. Accordingly, the image resolution correction apparatus 10 previously calculates the resolution decrease rate T indicating the extent of the spatial property degradation of the image calculated for each PTZ value of the dome camera, and stores it in matching with the PTZ value. The calculation of the resolution lowering rate T and the generation of the look-up table will be described below.

FIGS. 2 to 5 are diagrams illustrating a method of generating a lookup table for resolution correction of a dome camera according to an exemplary embodiment of the present invention. 6 is a flowchart illustrating a method of generating a lookup table for resolution correction according to an embodiment of the present invention. Hereinafter, description will be made with reference to FIGS. 2 to 6 together.

First, the camera captures a resolution chart under the condition that there is no dome cover (S601), and the camera captures the resolution chart under the condition that the dome cover is present (S602).

Resolution chart Before shooting, divide the area of the dome cover. Each divided dome cover area corresponds to the position of the dome camera according to the PTZ value of the dome camera. The PTZ value is set in the dome camera, and the dome camera takes a resolution chart with the determined angle.

When the dome camera shoots at the maximum zoom, the resolution degradation of the surveillance image due to the dome cover occurs the greatest. Hereinafter, a procedure for setting the dome camera to the maximum zooming state and generating a lookup table will be described in order to improve the resolution. However, the present invention is not limited to this, and it goes without saying that the lookup table can be generated by adjusting the pan / tilt (PT) value for each zoom magnification of the dome camera.

Referring to FIG. 2, if the area of the dome cover area occupied by the camera angle determined by the specific pan / tilt value of the dome camera in the maximum zoom state is defined as A, the total area A _total of the transparent dome cover is expressed by the following equation (1) Respectively. The image taken by a dome camera set to a specific pan / tilt value is S (k).

.....(1)

.....(One)

Referring to FIG. 3 (a), the image obtained by photographing the resolution chart at the pan tilt zoom value set by the dome camera under the condition of no dome cover is referred to as S_H (k). S_H (k) is an m x n pixel image. The image S_H (k) is an image without resolution degradation obtained without being affected by the dome cover.

S_H (1), S_H (2), ..., S_H (r)

Referring to FIG. 3 (b), the image obtained by photographing the resolution chart at the pan tilt zoom value set by the dome camera under the condition of the dome cover is referred to as S_L (k). S_L (k) is an m x n pixel image. The image S_L (k) is an image whose resolution is degraded by the refractive index and transmittance of the dome cover.

S_L (1), S_L (2), ..., S_L (r)

Next, DCT conversion is performed on the mxn pixel image S_H (k) acquired in the dome-free condition (S603), and DCT conversion is performed on the mxn pixel image S_L (k) acquired in the dome- .

Referring to FIG. 4, DCT transform is performed by dividing mxn pixel images S_H (k) and S_L (k) into 8x8 blocks, and performing discrete cosine transform for transforming time domain to frequency domain for each DCT block. .

DCT conversion is performed on the DCT blocks of the image S_H (k) photographed under the condition of no dome cover. C_H (k) represents a DCT coefficient calculated for all DCT blocks of the image S_H (k).

C_H (k) = C_H_k (1) + C_H_k (2) + C_H_k (3), ...

The DCT coefficients (C_H_k (1), C_H_k (2), C_H_k (3), ...) of each DCT block are arranged in order of components whose spatial characteristics are strong.

DCT conversion is performed on the DCT blocks of the image S_L (k) photographed under the condition of the dome cover. C_L (k) represents a DCT coefficient calculated for all DCT blocks of the image S_L (k).

C_L (k) = C_L_k (1) + C_L_k (2) + C_L_k (3), ...

The DCT coefficients C_L_k (1), C_L_k (2), C_L_k (3), ...) of each DCT block are arranged in order of components having strong spatial characteristics of images.

Analyzing the spectrum of an image reveals that a specific part of the spectrum has a large value for each case when there are many spatial components, a large number of horizontal components, and a large number of diagonal components in the image. Each DCT coefficient represents a spectrum including information on a spatial property of an image, i.e., a horizontality component, a verticality component, and a diagonal component.

The DCT coefficients representative of the spatial characteristics of the image, such as the horizontality component, the verticality component, and the diagonal component, are selected, and the following calculation is performed on the selected DCT coefficients to determine the spatial characteristics of the DCT block .

The DCT coefficient of the horizontal component: C_h (n) = mean (abs (C1) + abs (C2) + abs (C3) + abs

The DCT coefficient of the vertical component: C_v (n) = mean (abs (C5) + abs (C6) + abs (C7) + abs (C8))

The DCT coefficient of the diagonal component: C_d (n) = mean (abs (C9) + abs (C10) + abs (C11) + abs

Here, abs represents an absolute value function, and mean represents an average function.

For each DCT block, DCT coefficients are arranged in order from the dominant component having the largest value among the horizontal component, the vertical component, and the diagonal component in order to grasp the spatial characteristic. Table 1 below shows a result of comparing spatial property values calculated for DCT coefficients of a plurality of DCT blocks constituting the image S (k).

DCT block C_h C_v CD Spatial characteristic One 10 20 30 d-v-h 2 20 10 30 d-h-v 3 20 30 10 v-h-d ... ... ... ... ...

Referring to Table 1, the DCT coefficients of the first DCT block are arranged in the order of the diagonal component, the vertical component, and the horizontal component, because the first DCT block has strong spatial characteristics in the order of the diagonal component, the vertical component, and the horizontal component. The DCT coefficients of the second DCT block are arranged in the order of the diagonal component, the horizontal component, and the vertical component in order of the diagonal component, the horizontal component, and the vertical component. The third DCT block arranges the DCT coefficients in the order of the vertical component, the horizontal component, and the diagonal component, because the spatial components are strong in the order of the vertical component, the horizontal component, and the diagonal component.

Next, the DCT coefficient C_H (k) of the image S_H (k) photographed under the no-dome cover condition is compared with the DCT coefficient C_L (k) of the image S_L (k) photographed under the dome- The resolution lowering rate is calculated (S605).

Specifically, the DCT coefficient (C_H (k)) corresponding to the component having the largest value among the horizontal component, the vertical component, and the diagonal component of each DCT block under the condition without the dome cover decreases in the condition where the dome cover is present The resolution lowering rate T can be calculated by the following equation (2).

T (k) = C_L (k) / C_H (k) (2)

Here, T (k) consists of m x n DCT coefficients.

For example, referring to Table 1, T (k) indicates how much the diagonal components of the first DCT block and the second DCT block decrease under the dome-covered condition under the dome-free condition, And how much the vertical component of the block is reduced under the dome-covered condition.

Next, the resolution decrease rate T is matched with the pan / tilt / zoom value p, t, z and stored in the memory, thereby creating a lookup table for image resolution correction (S606). 5 is an illustration of a look-up table according to an embodiment of the present invention. Knowing the PTZ value of the dome camera, the spatial property of the input image from the lookup table and the degree of reduction of the spatial characteristic due to the dome cover can be known.

FIG. 7 is a flowchart illustrating a method of correcting the resolution of a surveillance image according to an embodiment of the present invention.

The image resolution correction apparatus of the present invention includes a dome camera having a transparent dome cover and PTZ controlled.

The image resolution correction device captures an image using a dome camera controlled by a set PTZ value (S701). The photographed image is an image formed by light passing through a specific area of the dome cover determined by the camera angle determined by the PTZ value.

The image resolution correction device performs DCT conversion of the photographed image (S703). The image data is converted into DCT coefficients by DCT conversion.

The image resolution correction apparatus corrects the DCT coefficient based on the spatial property of the image itself (S705). Specifically, coefficients of the high frequency region are selectively corrected by increasing the DCT coefficients by reflecting the resolution lowering rate obtained from the lookup table in the DCT coefficients. The lookup table is information for matching the PTZ value and the resolution lowering rate for image resolution correction. A method of generating a lookup table for correcting the resolution of an image has been described above, and thus a detailed description thereof will be omitted. The image resolution correcting device obtains the resolution lowering rate matched with the PTZ value at the time of image photographing from the lookup table and multiplies the inverse number by the DCT coefficient.

The image resolution correction device quantizes the corrected DCT coefficients (S707).

The image resolution correction apparatus codes the quantized coefficients (S709). The encoding may be performed using Run Length Coding (RLC) or Variable Length Coding (VLC).

For example, the current input image captured by the dome camera set to the first PTZ value is divided into a plurality of DCT blocks and subjected to DCT conversion. At this time, from the look-up table, it is possible to know the spatial property of the first DCT block and the degree of reduction (resolution decrease rate) of the spatial property by the dome cover. The DCT coefficient of the first DCT block is multiplied by the inverse of the resolution lowering rate to correct the DCT coefficient. The quantized and compressed process is performed on the corrected DCT coefficients.

The resolution correction method of the present invention can be applied not only to a system for selectively increasing the resolution of a surveillance image by image spectrum analysis but also to a system for reducing unnecessary portion resolution.

While the present invention has been described with reference to exemplary embodiments, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

Claims (7)

A photographing unit having a transparent dome cover and photographing a subject in a direction determined by a current pan tilt zoom value corresponding to one of the plurality of divided regions of the dome cover;
A discrete cosine transform unit for dividing an input image photographed with the current PTZ value into blocks of a predetermined size and calculating discrete cosine transform (DCT) coefficients for each block obtained by performing discrete cosine transform on image data of each block;
The first DCT coefficient for each block of the first image photographed with the pan tilt zoom value corresponding to each of the divided regions of the dome cover under the condition that the dome cover is not present and the first DCT coefficient for each of the divided regions of the dome cover A horizontal component of the first DCT coefficient and a vertical component of the first DCT coefficient for the corresponding block of the first image and the second image, which are calculated on the basis of a second DCT coefficient for each block of the second image photographed with the corresponding pan tilt zoom value, And a look-up table which stores a resolution lowering rate indicating a degree of reduction of a component having a largest value among the diagonal components in the second DCT coefficient, and stores the matching ratio with a pan tilt zoom value; And
And a resolution corrector for multiplying the block-by-block DCT coefficient of the input image by an inverse of a resolution reduction rate matched with the current pan tilt zoom value extracted from the look-up table. delete delete delete delete delete delete

Images