KR100290619B1 - Device and method for video coding by using selective quantizing step size - Google Patents

Abstract

PURPOSE: A video coding device by using a selective quantizing step size and a method thereof are provided to selectively apply a coding characteristic corresponding to a quantizing step for a certain portion of a video, thereby differentiating the screen quality of the portion from the other portions. CONSTITUTION: A video coding device by using a selective quantizing step size includes a coding characteristic allocation part(7) for receiving first to third coding characteristics from an editor with respect to respective NxN blocks divided from a first frame of a video, a coding characteristic memorizing part(8) for storing the characteristics by 2 bit characteristic values, a coding characteristic judging part(9) for searching a video via a motion estimation part(1) which finds out a most similar block to an estimation subject block from a previous frame, and applying a coding characteristic value having a highest distribution in a previous block as a coding characteristic of the current coding subject, and a variable step quantizing part(10) for carrying out the quantizing steps corresponding to the coding characteristic values by using an offset value input from the editor.

Description

Image Coding Apparatus and Method Using Selective Quantization Step Size

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to an encoding method for editing an image in a digital image storage device for digital video display (hereinafter referred to as DVD), and more particularly, to an encoding method in which an editor can make different reconstruction quality of a specific portion of an image. will be.

Currently, encoding for DVD image editing uses MPEG-2 image encoding, which will be described with reference to the accompanying drawings.

In the MPEG-2 video encoding method, one frame of an image is divided into 16 × 16 blocks, and a block matching method is used for each block, and the motion estimation unit 1 finds the most similar block in all frames. The motion compensation is performed through the motion compensator 2.

At this time, a difference occurs between the original image and the motion compensated image, and a transformation is performed by a two-dimensional discrete cosine transformer (DCT) 3 for the compensation error, which is input to the quantizer 4.

The quantizer 4 performs quantization using a quantization step determined for the two-dimensional DCT coefficients obtained using the two-dimensional DCT 3, and uses the variable length coder (VLC) 5 to quantize the quantized two-dimensional DCT coefficients. Encode

The value output through the quantizer 4 is inversely transformed through the two-dimensional inverse transformer (IDCT) 6 and then used as a reference value when reconstructing the next image.

However, the above coding method uses the same quantization step size for all parts of an image, and thus, when an editor wants to change the reconstruction quality of a specific part of the image, the reconstruction quality of the entire image is changed.

Disclosure of Invention The present invention aims at enabling quantization by varying the quantization step size according to the selection of an image editor, in view of the conventional problems as described above.

That is, the editor selects the fine encoder and the coarse encoder, and after that, the encoder determines based on this, so that the image quality can be controlled by changing the quantization step size.

1 is a block diagram showing a general MPEG-2 video encoder structure.

2 is a block diagram showing an MPEG-2 video encoder structure according to the present invention.

<Description of the reference numerals for the main parts of the drawings>

1: motion estimation unit 2: motion compensation unit

3: two-dimensional discrete cosine converter 5: variable length encoder

6: two-dimensional discrete cosine inverse transformer 7: coding characteristics providing unit

8: encoding characteristic memory unit 9: encoding characteristic determination unit

10: variable step quantization unit

In order to achieve the above object, an encoding apparatus implemented in the present invention includes an editor so that an editor may selectively encode a reconstruction quality for a specific object of an image;

An encoding characteristic assigning unit which gives an editor three encoding characteristics of the first characteristic, the second characteristic, and the third characteristic for each N × N block divided from the first frame of the image;

An encoding characteristic memory for storing each characteristic provided through the encoding characteristic providing unit using two-bit characteristic values;

An encoding characteristic determination unit for determining an encoding characteristic of the image and assigning the encoding characteristic to a block that is currently encoding target when the image is found through a motion estimation unit that finds the block most similar to the estimation target block from the previous frame;

A quantization step size is set differently according to a specific value provided by the encoding characteristic providing unit, and when the encoding characteristic value is applied to the current image through the encoding characteristic determination unit, the variable performs quantization in a quantization step corresponding to the encoding characteristic value. It is characterized by including a step quantization unit.

In addition, in order to achieve the above object, the encoding method proposed by the present invention includes the steps of an editor assigning encoding characteristics to a block of N × N partitioned in an encoding scheme;

Determining a coding characteristic of a motion compensation block obtained through motion estimation;

Including performing a quantization using a plurality of different quantization step sizes;

The editor selectively quantizes the reconstructed picture quality of a specific part of the image by applying different quantization step sizes.

The above objects, features, and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 2 is a block diagram of an encoding apparatus to which the encoding method of the present invention is applied. In the encoding characteristic granting unit 7, an editor displays an F (Fine) for each 16x16 block divided from the first frame of the video to be edited. Three coding characteristics, N (Normal) and C (Coarse), are given.

The assigned encoding characteristic is stored in the encoding characteristic memory 8 by using the characteristic value of two bits, and used to give the encoding characteristic for the blocks of the next frame.

The motion estimation unit 1 then finds the block most similar to the estimation target block from the previous frame.

The found block is used by the encoding characteristic determination unit 9 to determine the encoding characteristics of the current encoding target block from the distribution of the encoding characteristics of all frames.

The encoding characteristic is judged from the distribution of the encoding characteristic values in the block found from the previous frame through motion estimation. That is, the encoding characteristic value having the largest distribution in the block is given as the encoding characteristic of the block currently being encoded.

The encoding characteristics assigned to the blocks of the current frame through the above process are stored in the encoding characteristic memory 8 and used to determine the encoding characteristics of the next frame.

The variable step quantization unit 10 performs quantization on the compensation error resulting from the motion compensation using the corresponding quantization step from the newly-encoded encoding characteristic.

In detail, the encoding method of the present invention allows the editor to input the quantization step sizes for the three encoding characteristics. The editor inputs the offset value for the quantization step before editing, so that the encoding characteristic is N. In the case of F, the quantization step size is used, and in the case of F, the value obtained by subtracting the offset value from the quantization step size of N is used. In case of C, the quantization is performed using the value obtained by adding the offset value from the quantization step size of N. .

As described in detail above, by using the encoding method of the present invention, an editor can selectively assign encoding characteristics corresponding to quantization steps for a specific portion of an image to be edited, which restores a specific portion during image restoration. The picture quality can be different from other parts of the image.

Such a feature has the advantage of providing a function of highlighting or blurring a specific object when editing a DVD image.

In addition, a preferred embodiment of the present invention is disclosed for the purpose of illustration, those skilled in the art will be able to make various modifications, changes, additions, etc. within the spirit and scope of the present invention, such modifications and modifications belong to the following claims You will have to look.

Claims (6)

In the compression encoding apparatus for video data, Allowing the editor to selectively encode the reconstruction quality differently for a specific object of the image; An encoding characteristic assigning unit which receives three encoding characteristics of the first, second and third characteristics from the editor for each N × N block divided from the first frame of the image; An encoding characteristic memory for storing each characteristic provided through the encoding characteristic providing unit using two-bit characteristic values; After searching an image through a motion estimation unit that finds the block most similar to the estimation target block from the previous frame, the encoding characteristic having the largest distribution in the searched previous block is assigned to the encoding characteristic of the block to be currently encoded. An encoding characteristic determination unit; After receiving the offset value for the quantization step from the editor, if the encoding characteristic value is given to the current image through the encoding characteristic determination unit, quantization is performed in the quantization step corresponding to each encoding characteristic value using the offset value An image encoding apparatus using a selective quantization step size, comprising: a variable step quantization unit. The method of claim 1, In each of the characteristics provided by the encoding characteristic provision unit, a first characteristic is a characteristic to be given to an image to be encoded with a higher image quality than normal image quality. The second characteristic is a characteristic to be given to an image to be encoded with normal picture quality. The third characteristic is an image encoding apparatus using the selective quantization step size, characterized in that the characteristic to give to the image (Coarse) to be encoded in a lower quality than the image quality of the calculated. The method of claim 1, When the quantization step is determined by the variable step quantization unit, If the encoding characteristic determined by the encoding characteristic determination unit is F, quantization is performed in a quantization step obtained by subtracting the offset value from a preset quantization step. If the encoding characteristic determined by the encoding characteristic determination unit is N, quantization is performed in a predetermined quantization step. And if the encoding characteristic determined by the encoding characteristic determination unit is C, quantization is performed in a quantization step that adds the offset value in a preset quantization step. In the compression encoding method of video data, A first step of receiving encoding characteristics from an editor for an N × N block divided from a first frame of an image to be encoded; After searching an image through a motion estimation unit that finds the block most similar to the estimation target block from the previous frame, the encoding characteristic value having the largest distribution in the searched previous block is determined as the encoding characteristic of the block to be currently encoded. A second process; A third process of performing quantization by applying a plurality of different quantization step sizes according to the determined encoding characteristics using the offset value input from the editor; An image encoding method using an optional quantization step size, characterized by allowing an editor to selectively encode a reconstruction quality for a specific object of an image. The method of claim 4, wherein As the encoding characteristic received in the first step, the encoding characteristic may be provided to an image to be encoded with higher quality than normal image quality; A property of giving an image to be encoded with normal picture quality (Normal); A video encoding method using a selective quantization step size, characterized in that it has a characteristic of giving to a video to be encoded with a lower quality than normal picture quality. The method of claim 4, wherein In the quantization in the third process, If the encoding characteristic determined by the encoding characteristic determination unit is F, quantization is performed in a quantization step obtained by subtracting the offset value from a preset quantization step. If the encoding characteristic determined by the encoding characteristic determination unit is N, quantization is performed in a predetermined quantization step. And if the encoding characteristic determined by the encoding characteristic determination unit is C, quantization is performed in a quantization step that adds the offset value in a preset quantization step.

Images