KR20110062448A - Apparatus and method for improving quality of preview image - Google Patents

Abstract

PURPOSE: A device and method for improving the quality of preview image are provided to improve an aliasing phenomenon and resolution deterioration generated form image preview image if sampling rate must be set high. CONSTITUTION: A sub sampling unit(200) performs a sub sampling selecting a basic barrier pattern per every sub window about n number frame of an input image. An image signal processing unit(230) generates preview image by using each frame which is sub sampled. A correlation multi sampling unit(210) removes a noise about a basic bayer pattern selected form each frame which is sub sampled.

Description

Apparatus and method for improving the quality of preview image {APPARATUS AND METHOD FOR IMPROVING QUALITY OF PREVIEW IMAGE}

The present invention relates to an apparatus and method for improving image quality, and more particularly, to an apparatus and method for improving image quality of a preview image.

In general, an image sensor such as a camera, a camcorder, a webcam, or the like is equipped with an image sensor. The image sensor captures an image deployed in a field of view, converts it into an electrical signal, and converts the converted image signal into a digital signal. The image signal output through the image sensor is red, green, and blue color image data of three colors, and the image data is displayed on the display unit. An image sensor applied to such an imaging device may perform a subsampling mode operation of outputting a full frame image signal by lowering the resolution. The sub-sampling mode is performed for fast signal processing in a step that does not need to be displayed in high resolution, such as a preview step of confirming in advance before capturing an image.

In such a subsampling mode, the resolution is lowered by selecting only pixels of specific rows and columns spaced at regular intervals for subsampling. For example, referring to FIG. 1, FIG. 1 illustrates a frame image composed of a plurality of Bayer patterns 140, and a subsampling method for generating a conventional preview image is as follows.

As shown in FIG. 1, when sampling at 1/4 resolution in the horizontal and vertical directions at n × m pixels, the image is divided into four sub-windows 100, 110, 120, and 130, one for each of the sub-windows 100, 110, 120, and 130. Bayer patterns 140, 150, 160 and 170 are selected. This image is a Bayer pattern in which two (R) and G (Green) two-color patterns are arranged in one row, and two (G) and Green (B) blue patterns are repeatedly arranged in another row. Have As such, only one Bayer pattern intersecting in the horizontal and vertical directions is selected and the other is removed to reduce the resolution, thereby reducing the amount of data to be processed. In the case of such a subsampling, only the same pixel data is always selected and used for all frames. As a result, the spatial frequency sampled by the image sensor is always the same.

In the conventional subsampling mode, only pixels of a specific row and column are used for subsampling, and the amount of data processed can be reduced, but there are many pixels that are discarded without being used. The aliasing problem that appears is large. Also, this is not a problem when the subsampling is not performed or the sampling rate is low. However, when the sampling rate is high, the spatial frequency is lowered, which causes serious staircase and resolution degradation in the preview image. Accordingly, when a high sampling rate needs to be set, a method for improving a step phenomenon and a resolution degradation occurring in the preview image is required.

Accordingly, the present invention provides an apparatus and method for improving the quality of a preview image in the sub-sampling mode.

According to an aspect of the present invention, there is provided a method for improving the quality of a preview image in a video device, the method comprising: performing subsampling to select a basic Bayer pattern for each subwindow for an nth frame of an input image; Performing subsampling for selecting a basic Bayer pattern based on a subsampling starting point shifted by a predetermined integer with respect to the first subsampling starting point for each subwindow for the n + 1th frame of the input image; and the subsampling And generating a preview image by using each of the frames.

In addition, the present invention, in the video device for improving the quality of the preview image, performing a subsampling to select a basic Bayer pattern for each sub-window for the n-th frame of the input image, the n + 1th frame of the input image A subsampling unit for performing subsampling for selecting a basic Bayer pattern based on a subsampling starting point shifted by a predetermined integer with respect to the first subsampling starting point for each subwindow, and previewing each subsampled frame And an image signal processor for generating an image.

According to the present invention, there is an advantage that the aliasing problem is improved when the sampling rate is to be set high. Accordingly, the image quality of the preview image can be improved.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

The present invention proposes a method for improving the image quality of a preview image. To this end, when an image is input in a frame unit, the present invention performs subsampling to select a Bayer pattern having a different subsampling starting point for each input frame, and adjusts the correlation between the frames of each subsampled to obtain a preview image. Create and display the generated preview image sequentially. By doing so, it is possible to improve the aliasing phenomenon and the resolution degradation occurring in the preview image when the sampling rate is to be set high.

Referring to FIG. 2, a method for improving the quality of a preview image in an image device in which the above function is implemented is described. 2 is a block diagram illustrating an internal block of a video device for improving preview image quality according to an embodiment of the present invention. Here, the video device may be a camera, a camcorder, a webcam, or the like.

First, when an image is input from the image device, the image input for displaying the preview image is transmitted to the sub-sampler 200 in units of frames. The sub-sampling unit 200 performs subsampling by selecting only the pixel data necessary for the input image and deleting the rest for the preview image.

According to an exemplary embodiment of the present invention, the sub-sampling unit 200 selects one basic Bayer pattern for each sub-window based on the a-row and b-column, that is, the position (a, b) in the nth frame. In the n + 1th frame, one basic Bayer pattern is selected for each sub-window based on the position (a + k1, b + k2) shifted by k1 columns and k2 rows based on the (a, b) positions. In this case, k1 and k2 are integers and may be the same value.

Subsequently, the subsampling process may be repeated on the n + 2 th frame or less based on the same position as the n th frame and the n + 1 th frame. Alternatively, in the n + 2th frame, subsampling may be performed by constantly shifting the starting point, such as with respect to the position (a + k3, b + k4). As such, the sub-sampling unit 200 performs subsampling at a predetermined sampling rate x. Here, the Bayer pattern refers to a pattern in which two patterns of R (Red) and G (Green) are arranged in one row and two patterns of G (Green) and B (Blue) in another row. In the Bayer pattern, the basic Bayer pattern means four pixel data in which first rows are arranged with G and R pixels, and second rows are arranged with B and G pixels.

The pixel value of each subsampled image includes noise due to internal factors or external factors, and when the preview image is generated using the pixel value including the noise, it may be different from the actual captured image. It must be calibrated. Accordingly, the correlated double sampling unit (CDS) 210 removes noise with respect to the subsampled image, that is, the selected basic Bayer pattern. As described above, flicker may be improved through noise reduction from the inter-frame relationship.

Each basic Bayer pattern of the noise-removed image is converted into digital data by an analog-to-digital converter (ADC) 220. Then, the data converted into digital data is input to the image signal processor 230, and the image signal processor 230 performs correlation between the frame and the pixel based on the initial subsampling starting point for each frame whose subsampling starting point is shifted. Create a preview image by fitting.

When the correlated preview image is input to the display unit 240, the preview image is displayed through the display unit 240. Accordingly, the preview viewer of the display unit 240 displays the preview image on a frame-by-frame basis, and each frame has a parallax of 1 / frame ratio. In general, however, since the preview image is displayed very quickly, the user does not feel parallax due to an optical illusion, but by selecting pixel data at different positions according to a frame, unlike a subsampling method using only pixel data at fixed positions. The staircase phenomenon and the degradation of the resolution occurring in the preview image are improved. In addition, since the subsampling start point is shifted and subsampled, the subsampled frame can be used to generate the preview image immediately without having to temporarily store it in a buffer to process the subsampled frame, thereby eliminating the need for hardware change.

An operation in the video device having the above configuration will be described with reference to FIG. 3. 3 is a flowchart illustrating an operation of an imaging apparatus according to an exemplary embodiment of the present invention. Hereinafter, to help understand the embodiment of the present invention will be described with reference to FIGS.

Referring to FIG. 3, when an image is input in frame units in operation 300, the image apparatus sets n to 1, which indicates the number of frames, in step 305, and sets k, which represents a shift integer for shifting a subsampling starting point, into rows and columns, to 0. The sampling rate (x). Then, in step 310, subsampling is performed from the position (a + k, b + k) of the nth frame.

To describe this in detail, referring to FIG. 4, FIG. 4 has a pattern in which first and second pixels G and R pixels are alternately arranged, and the second row has a pattern in which B and G pixels are alternately arranged. Accordingly, an odd row has a repeating pattern of a first row, and an even row illustrates a frame of an input image in which n × m pixels in a two-dimensional matrix form having a repeating pattern of a second row are arranged.

Specifically, in FIG. 4, one frame is divided into a plurality of sub-windows 445, 450, 455, 460,... According to horizontal 1/4 and vertical 1/4 sub-sampling in one frame of the input image. The case where one basic Bayer pattern 400, 405, 410, 415, ... is selected for each of the sub windows 445, 450, 455, 460, ... is illustrated. That is, the case of subsampling at 1/4 resolution is illustrated, and the division of the sub window is determined according to the 1 / n resolution. As described above, in one frame, pixel data selection is performed in units of basic Bayer patterns.

After performing subsampling to select one basic Bayer pattern for each subwindow in the same frame as in FIG. 4, if the next frame is input in step 315, the number of frames is increased by one as n = n + 1 in step 320. And k = k + y to shift the subsampling starting point, where y is an integer. Then, it is determined whether k is equal to or greater than the sampling rate x, and if the determination result is not equal to or greater than the sampling rate x, the process returns to step 310 and repeats the above process, but k is greater than or equal to the sampling rate x. If k = 0 is initialized, the process returns to step 310. Then, if the next frame is not input in step 315, the process proceeds to step 335 to generate and display a preview image using the selected basic Bayer pattern.

Referring to FIG. 5 to describe this in detail, FIG. 5 illustrates an n + 1th frame if FIG. 4 illustrates an nth frame. Assuming that the subsampling starting point is the position of the basic bayer pattern as indicated by 400 in the nth frame, the subsampling starting point in the n + 1th frame is the basic Bayer pattern as indicated by 500 shifted by a predetermined integer. Becomes the position of. Accordingly, as shown in 500, one basic Bayer pattern 500, 505, 510, 515 is selected for each sub-window starting with the position of the basic Bayer pattern.

Therefore, in the regularity of the basic Bayer pattern selection, if the sub-sampling starting point is changed twice as shown in Figs. 4 and 5, the subsampling operation and n + 1 as in the nth frame to generate the preview image The subsampling operation as in the first frame is alternately repeated.

On the contrary, in the case where the subsampling starting point is changed for each frame, the subsampling starting point is sequentially shifted by a predetermined integer. Accordingly, the shift constant k in step 320 is added by a predetermined integer y. In addition, in the case of shifting by the same k in the row and column directions, respectively, based on the first sub-sampling starting points a and b in step 310, the shift k in the row and column directions may be different. At this time, the shift constant k is increased until it is not larger than the sampling ratio x so that the subsampling starting point can be shifted by the shift constant. However, when the shifting constant k is higher than the sampling ratio x, the shift constant k is out of the sub window. It is preferable to reset the subsampling start point to 0 so that the subsampling start point is the first subsampling start point.

As described above, in FIG. 3, the subsampling operation in the n, n + 1, n + 2, and the like frames having different subsampling starting points is explained. Each of the basic Bayer patterns selected in each of the frames is arranged. Referring to the frame with respect to the time axis, it is as shown in FIG. 6. As shown in FIG. 6, in the nth frame, the position of the basic Bayer pattern indicated by 400 becomes the first subsampling starting point, and in the n + 1th frame, the position of the Basic Bayer pattern indicated by 500 corresponds to the subsampling starting point In the n + 2th frame, the position of the basic Bayer pattern indicated by 600 becomes the sub sampling start point. 6 illustrates a spatial frequency viewed from the time axis according to an embodiment of the present invention.

As shown in FIG. 6, the video apparatus according to the exemplary embodiment of the present invention selects a basic Bayer pattern for each subwindow based on a position shifted by a predetermined integer based on the first subsampling starting point according to the frame input. Do this. Then, for each subsampled frame, a preview image is generated by performing a correlation operation based on an initial subsampling start point and removing noise.

1 is a diagram illustrating a subsampling method for generating a conventional preview image;

2 is a block diagram illustrating an internal block of a video device for improving preview image quality according to an embodiment of the present invention;

3 is an operation flowchart of an imaging device according to an embodiment of the present invention;

4 to 6 are exemplary diagrams for explaining the sub-sampling method according to the present invention.

Claims (14)

In the method for improving the quality of the preview image in the video device, Performing subsampling to select a basic Bayer pattern for each subwindow for the nth frame of the input image; Performing subsampling for selecting a basic Bayer pattern with respect to the n + 1th frame of the input image based on a subsampling starting point shifted by a predetermined integer with respect to the first subsampling starting point for each subwindow; And generating a preview image by using each of the subsampled frames. The method of claim 1, wherein the basic Bayer pattern, A method for improving the image quality of a preview image, characterized in that the pattern is composed of four pixel data in which red and green pixel data are arranged in one row and green and blue pixel data are arranged in another row. The method of claim 1, Performing subsampling based on the first subsampling starting point or performing subsampling based on the shifted subsampling starting point for each frame input after the n + 1th frame of the input image; And improving the quality of the preview image. The method of claim 1, Performing subsampling for selecting a basic Bayer pattern on the basis of a subsampling starting point shifted by a predetermined integer based on a previous subsampling starting point each time a frame is input after an n + 1th frame of the input image; The method for improving the quality of the preview image, characterized in that it further comprises. The image quality of the preview image of claim 4, further comprising: setting the sub sampling start point at the next frame input as the first sub sampling start point when the shifted sub sampling start point is equal to or greater than the sampling rate. How to. The method of claim 1, wherein the generating of the preview image using each subsampled frame comprises: Removing noise for the selected basic Bayer pattern of each of the subsampled frames; And digitally converting the noise-free basic Bayer pattern to generate the preview image. The method of claim 1, And correlating the subsampled frames with respect to each of the subsampled frames based on the initial subsampling starting point. In the video device for improving the quality of the preview video, Performing subsampling to select a basic Bayer pattern for each subwindow for the nth frame of the input image, and shifting a predetermined integer based on a starting point of the first subsampling for each subwindow for the n + 1th frame of the input image A subsampling unit which performs subsampling to select a basic Bayer pattern based on a subsampling starting point; And an image signal processor configured to generate a preview image using each of the subsampled frames. The method of claim 8, wherein the basic Bayer pattern, Image device for improving the quality of the preview image, characterized in that the pattern is composed of four pixel data arranged in one row of red, green color pixel data and green, blue pixel data in another row . The method of claim 8, wherein the sub-sampling unit, And performing subsampling based on the first subsampling starting point or performing subsampling based on the shifted subsampling starting point for each frame input after the n + 1th frame of the input image. Imaging device to improve the quality of preview video. The method of claim 8, wherein the sub-sampling unit, Whenever a frame is input after the n + 1 th frame of the input image, subsampling is performed by selecting a basic Bayer pattern based on a subsampling starting point shifted by the predetermined integer based on a previous subsampling starting point. A video device for improving the quality of preview video. The method of claim 8, wherein the sub-sampling unit, And when the shifted subsampling start point is equal to or greater than a sampling rate, a subsampling start point at the next frame input as an initial subsampling start point. The method of claim 8, wherein the image signal processing unit, And generating the preview image by correlating each of the subsampled frames based on the initial subsampling starting point. The method of claim 8, A correlated double sampling unit for removing noise with respect to the selected basic Bayer pattern of each of the subsampled frames; And an analog-to-digital converter configured to digitally convert the noise-free basic Bayer pattern.

Images