KR20180092652A - Method for adaptive scene change detection based on resolution and apparatus for the same - Google Patents

Abstract

An adaptive scene change detection method and apparatus are disclosed. The adaptive scene change detection method includes a step of extracting a luminance component in each of a current frame and a previous frame of the current frame, a step of performing binary encoding on the luminance component based on a block predetermined according to a resolution in each of the current frame and the previous frame, a step of comparing corresponding blocks with the luminance components of the current frame and the previous frame that are binary encoded and determining a scene change candidate block of the current frame, and a step of performing directional distribution analysis on the scene change candidate block and selecting a scene change block. Therefore, the scene change detection can be adaptively performed according to the resolution.

Description

[0001] The present invention relates to an adaptive scene change detection method and apparatus,

The present invention relates to an adaptive scene change detection method and apparatus, and more particularly to an adaptive scene change detection method and apparatus that adaptively determines a size of a scene change block according to a resolution, To a scene change detection method and apparatus.

LCD (Liquid Crystal Display), which is widely used in recent years, has a hold-type display characteristic which keeps a long screen per each frame and affects the next frame. Therefore, a so-called motion blur effect in which a residual image superimposed on a previous frame appears in a new frame is caused, which causes image quality deterioration.

As a solution of this phenomenon, Frame Rate Up Conversion (FRUC) has been proposed, and the most commonly used frame rate enhancement technique is a motion compensated frame rate enhancement technique considering motion. The motion compensation frame rate enhancement technique is a technique of generating and adding an interpolation frame between adjacent frames through motion prediction and compensation. However, when a scene changes suddenly when an object appears or changes into a new scene, an interpolation frame is generated There is a problem that is difficult to detect and a technique of detecting a scene change is required.

In addition, it is important to detect transitions in video indexes that make it easy for users to find the information they need from a vast amount of video.

There are various techniques of the existing scene change research. For example, various feature information such as pixel, histogram, edge component, texture, motion, and DCT information of a frame is used to measure a change between frames through comparison of neighboring frames.

The basic principle of the method of detecting the scene change through the comparison of adjacent frames is to detect a scene change when the feature value difference between neighboring frames exceeds a predetermined threshold value, It is important to choose an appropriate threshold value. However, since most of existing technologies mainly use fixed threshold values, a higher quality image can degrade performance, and it is difficult to accurately detect a scene depending on resolution.

Also, there is a problem that it is difficult to detect a scene according to the quality of a video because a fixed block size is used in comparison between blocks.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide an adaptive scene change detection method.

It is another object of the present invention to provide an adaptive scene change detection apparatus.

According to an aspect of the present invention, there is provided an adaptive scene change detection method.

Here, the adaptive scene change detection method includes: extracting a luminance component in each of a current frame and a previous frame of a current frame; performing a binary coding on a luminance component based on a predetermined block according to a resolution in each of a current frame and a previous frame Determining a scene change candidate block of a current frame by comparing corresponding blocks with respect to a luminance component of a current frame and a luminance component of a previous frame, performing a directional distribution analysis on the scene change candidate block, .

Here, after the step of extracting the luminance component, filtering may be performed on each of the luminance components extracted from the current frame and the previous frame using a low-pass filter.

Here, the predetermined block according to the resolution may be determined to have a size proportional to the horizontal resolution of the current frame.

Here, the predetermined block according to the resolution is determined to have the size of BkSize x BkSize using the block width (BkSize) determined by the following equation,

Here, Video_reso_width may be the horizontal resolution of the current frame.

Here, the step of determining a scene change candidate block may include determining an average of pixel value differences between a block of a current frame and a block of a previous frame, with a first threshold value.

Here, the step of determining a scene change candidate block may include: XORing a pixel value between blocks of a current frame and a previous frame; comparing an average of XOR (exclusive OR, exclusive-OR) values with a second threshold value And a step of determining the number

Here, the first threshold value may be adaptively determined according to a size of a predetermined block.

Here, the first threshold value may increase linearly as the size of a predetermined block increases.

Here, after the step of determining the scene change candidate block, the step of dividing the determined scene change candidate block into quad trees and comparing each of the divided blocks with the block of the previous frame corresponding thereto to select scene change candidate blocks .

Here, after the step of determining the scene change candidate block, the step of reconstructing the scene change candidate block may exclude a block that can be predicted through motion estimation from the scene change candidate block.

According to another aspect of the present invention, there is provided an adaptive scene change detection apparatus.

Here, the adaptive scene change detection apparatus includes a processor for executing at least one instruction and a memory for storing at least one instruction.

Here, the processor extracts a luminance component in each of the current frame and the previous frame of the current frame, performs binary coding on the luminance component based on a predetermined block according to the resolution in the current frame and the previous frame, The scene change candidate block of the current frame is determined by comparing the current frame and the previous frame with the corresponding blocks of the predetermined block, and the scene change block is selected by performing directional distribution analysis on the scene change candidate block.

Here, the processor may filter the luminance components extracted from the current frame and the previous frame using a low-pass filter.

Here, the predetermined block according to the resolution may be determined to have a size proportional to the horizontal resolution of the current frame.

Here, the predetermined block according to the resolution is determined to have the size of BkSize x BkSize using the block width (BkSize) determined by the following equation,

Here, Video_reso_width may be the horizontal resolution of the current frame.

Here, the processor may determine a scene change candidate block by comparing an average of pixel value differences between a block of a current frame and a block of a previous frame, with a first threshold value.

Here, the processor may determine the scene change candidate block by performing an XOR operation on pixel values between blocks of a current frame and a previous frame, and comparing an average value of XOR (exclusive OR, exclusive-OR) have.

Here, the first threshold value may be adaptively determined according to a size of a predetermined block.

Here, the first threshold value may increase linearly as the size of a predetermined block increases.

Here, the processor may select a scene change candidate block by dividing the determined scene change candidate block into quad trees, and comparing each of the divided blocks with a block of a previous frame corresponding thereto.

Here, the processor can reconstruct a scene change candidate block by excluding blocks predictable through motion estimation from the scene change candidate blocks.

When the adaptive scene change detection method and apparatus according to the present invention as described above is used, it is possible to perform scene change detection adaptively according to the resolution.

Also, there is an advantage that accuracy of scene change detection can be improved in a high resolution image.

In addition, the performance time and signal-to-noise ratio (PSNR) can be improved as compared with the conventional method.

1 is a flowchart of an adaptive scene change detection method according to an embodiment of the present invention.
FIG. 2 is an exemplary diagram illustrating a process of comparing blocks to determine a scene change according to an exemplary embodiment of the present invention. Referring to FIG.
3 is a diagram illustrating an example of determining a scene change candidate block using an average pixel value according to an exemplary embodiment of the present invention.
4 is a diagram illustrating an example of determining a scene change candidate block using an XOR operation according to an embodiment of the present invention.
5 is a graph for determining an optimal first threshold value through a linear regression analysis according to an embodiment of the present invention.
FIG. 6 is a diagram illustrating a process of dividing a scene change candidate block according to an embodiment of the present invention.
7 is an exemplary diagram for explaining directional distribution analysis according to an embodiment of the present invention.
8 is a block diagram of an adaptive scene change detection apparatus according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, A, B, etc. may be used to describe various elements, but the elements should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.

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

1 is a flowchart of an adaptive scene change detection method according to an embodiment of the present invention.

Referring to FIG. 1, an adaptive scene change detection method includes a step S100 of extracting a luminance component in a current frame and a previous frame in a current frame, respectively, (S120) of determining a scene change candidate block of a current frame by comparing the corresponding blocks of the binary-coded current frame with a luminance component of a previous frame (S120), and performing a binary encoding on the scene change candidate block And performing a directional distribution analysis on the scene change block (S130).

Here, after the step of extracting the luminance component (S100), the step of filtering each luminance component extracted in the current frame and the previous frame using the low-pass filter may be further included.

Here, the step of filtering using a low-pass filter may have the effect of alleviating a sudden change between surrounding pixels.

Here, in the step of performing the binary encoding (S110) or the step of determining the scene change candidate block (S120), the scene detection performance may be greatly affected depending on the size of the block because it is performed on a block-by-block basis.

For example, when a block having 8 × 8 pixels is used for video in the CIF (Common Intermediate Format) format, about 1500 blocks can be generated in the CIF standard resolution (352 × 288). However, in the case of HD (High Definition) video, since the number of blocks is increased ten times or more as compared with the CIF format, the complexity of detecting scene change can be greatly increased. Therefore, it may be useful to predetermine the size of the block according to the resolution.

Therefore, the predetermined block according to the resolution can be determined to have a size proportional to the horizontal resolution of the current frame.

Here, the predetermined block according to the resolution can be determined to have the size of BkSize x BkSize using the block width (BkSize) determined by the following equation (1).

는 반올림으로 연산됨을 의미할 수 있다.Here, Video_reso_width can indicate the width of the frame,

May be computed as rounding.

When the size of the block is determined according to the resolution, when a frame sequence belonging to Class C is used, it has a block size of 16 × 16 because it has a resolution of 832 × 480. In this case, Compared with the block size of 32 × 32, the execution time of scene transition detection can be reduced by 1.98% and 1.14%, respectively.

The block size according to the representative resolution according to Equation (1) can be determined as shown in Table 1 below.

Here, the step of performing the binary encoding (S110) will be described in detail as follows.

)을 계산할 수 있다. 여기서, x와 y는 각 픽셀의 위치, N과 M은 블록 내 수직 및 수평 픽셀의 수를 의미할 수 있다. Referring to Equation (2), for each block in the current frame and the previous frame, a pixel average value (

) Can be calculated. Where x and y are the position of each pixel, and N and M are the number of vertical and horizontal pixels in the block.

)을 블록 내의 각 픽셀(p_i,j(x,y)) 과 비교하여 각 픽셀값이 픽셀 평균값보다 크다면 1로, 그렇지 않으면 0으로 결정함으로써, 각 픽셀값을 이진화할 수 있고, 이러한 과정을 현재 프레임과 이전 프레임 각 블록에 대해서 수행하여 이진 부호화를 수행할 수 있다.Referring to Equation (3), the pixel average value (

) Is compared with each pixel (p _{i, j} (x, y)) in the block and each pixel value is determined to be 1 if the pixel value is greater than the pixel average value, otherwise, each pixel value can be binarized, May be performed for each block of the current frame and the previous frame to perform binary coding.

Hereinafter, the step of determining a scene change candidate block (S120) can be described in detail with reference to the drawings.

FIG. 2 is an exemplary diagram illustrating a process of comparing blocks to determine a scene change according to an exemplary embodiment of the present invention. Referring to FIG.

Referring to FIG. 2, the previous frame 10 and the current frame 20 may be divided into predetermined blocks, respectively. At this time, the blocks located at the same position may be corresponding blocks for judging whether to switch the scene.

Therefore, in the previous frame 10, the first upper left block 11 and the upper left upper block 21 in the current frame 20 may be the corresponding blocks for judging scene change.

3 is a diagram illustrating an example of determining a scene change candidate block using an average pixel value according to an exemplary embodiment of the present invention.

Referring to FIG. 3, assuming that the block 11 of the previous frame and the block 21 of the current frame correspond to each other as shown in FIG. 2 and correspond to a comparison block for detecting a scene change, An average pixel value can be derived.

이 계산될 수 있다. 두 블록 간 평균 픽셀 값의 차분을 구하면, 0.34375가 도출될 수 있고, 이 차이값을 제1 임계값과 비교함으로써 장면 전환 블록이 될 수 있는지 여부를 결정할 수 있다.3, 0.5 is calculated as an average pixel value for the block 11 of the previous frame and an average pixel value is calculated for the block 21 of the current frame

Can be calculated. If the difference of the average pixel value between two blocks is obtained, 0.34375 can be derived and it can be determined whether or not it can become a scene change block by comparing this difference value with the first threshold value.

This can be summarized as follows.

는 이전 프레임의 해당 블록 평균 픽셀값이고,

는 현재 프레임의 해당 블록 평균 픽셀값이며,

는 비교되는 두 블록 간 평균 픽셀 값의 차분값일 수 있다.Referring to Equation 4,

Is the corresponding block average pixel value of the previous frame,

Is the corresponding block average pixel value of the current frame,

May be the difference value of the average pixel value between the two compared blocks.

The following equation is used to determine the scene change candidate block using the difference value.

)이 제1 임계값(T1) 보다 크면 해당 블록을 1로, 그렇지 않으면 0으로 지정함으로써 (i.j)에 위치한 블록(CB_i,j)이 장면 전환 후보 블록인지 여부를 표현할 수 있다.Referring to Equation (5), the difference value of the average pixel value between two blocks

) Is greater than the first threshold value T1, it is possible to express whether the block (CB _{i, j} ) located at (ij) is a scene change candidate block by designating the corresponding block as 1 or 0 otherwise.

Referring again to FIG. 1, the step of determining a scene change candidate block S120 includes a step of comparing an average of pixel value differences between blocks of a current frame and blocks of a previous frame with a first threshold value can do.

Here, determining the scene change candidate block only by the difference of the average pixel value may involve a large error. Therefore, it is necessary to consider an additional index to determine the scene change candidate block. Comparing with XOR operation can be helpful considering binary coding.

This will be described in detail below.

4 is a diagram illustrating an example of determining a scene change candidate block using an XOR operation according to an embodiment of the present invention.

Referring to FIG. 4, the scene change candidate block can be determined by XORing the block 11 of the previous frame and the block 21 of the current frame.

Specifically, an XOR block 22 in which an XOR operation is performed between blocks can be derived. The scene change candidate block can be determined by obtaining the binary-coded pixel value average of the derived XOR block 22.

In the binary coded block in Fig. 4, 0.5 can be derived as the pixel average value of the XOR block 22. [

The scene change candidate block can be determined more precisely by adding the process of comparing the pixel average value thus derived with the second threshold value.

This can be expressed as follows.

는 현재 프레임과 이전 프레임의 비교 블록 간 XOR 연산한 픽셀 평균값을 의미할 수 있고, 그밖의 부호는 앞에서의 수학식들에서 설명한 바와 동일할 수 있다.here,

May mean a pixel average value obtained by performing an XOR operation on a comparison block between a current frame and a previous frame, and the other symbols may be the same as described in the above-mentioned mathematical expressions.

By considering the XOR operation as described above, the scene change candidate block can be determined as follows.

)을 제2 임계값과 비교한 결과 모두가 제1임계값 및 제2 임계값보다 큰 경우에 해당 블록을 1로, 그렇지 않으면 0으로 지정함으로써 (i.j)에 위치한 블록(CB_i,j)이 장면 전환 후보 블록인지 여부를 표현할 수 있다.Referring to Equation (7), the comparison result with the first threshold value according to Equation (5) and the pixel average value XOR calculated

(CB _{i, j} ) positioned at (ij) by designating the corresponding block as 1 if all of the comparison result with the second threshold value is greater than the first threshold value and the second threshold value _, Whether or not the scene change candidate block is a scene change candidate block.

 Referring to FIG. 1 again, the step of determining a scene change candidate block (S120) comprises XORing (XORing) pixel values between blocks of the current frame and blocks of the previous frame, 2 < / RTI > threshold value.

On the other hand, when the first threshold value is used as a fixed value, the scene change can not be adaptively determined according to the characteristics of the image or the size of the block, so that the performance may be deteriorated.

It may be advantageous to apply an adaptive threshold value in view of this and can be explained with reference to the drawings hereinafter.

5 is a graph for determining an optimal first threshold value through a linear regression analysis according to an embodiment of the present invention.

Referring to FIG. 5, the first threshold value may be adaptively determined according to the block size.

)을 장면 전환 유무로 나누어 구함으로써 평균값을 구하고, 이 값을 바탕으로 선형 회귀 분석을 수행하여 최적의 제1 임계값을 결정할 수 있다.Specifically, when the block size is 8 × 8, 16 × 16, or 32 × 32, the difference value of the average pixel value between the two blocks

) Is divided by the presence or absence of the scene change, and the average value is obtained, and the optimal first threshold value can be determined by performing a linear regression analysis based on the average value.

As a result of the preceding regression analysis, the first threshold value T1 can be determined as follows.

The first threshold value T1 may be determined according to the number of times and the number of times the linear regression analysis is performed.

(8) and (9), the first threshold value can be adaptively determined according to the size of the predetermined block, and specifically, the first threshold value increases linearly as the size of the predetermined block increases .

In this manner, the scene change candidate block is selected and the scene change candidate block is selected. In this case, it may be advantageous to select the scene change candidate block in smaller block units than the scene change candidate block is used as it is.

For example, it may be advantageous to select the size of the scene change candidate block to be smaller if the scene change area is small, such as when local scene change occurs.

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

FIG. 6 is a diagram illustrating a process of dividing a scene change candidate block according to an embodiment of the present invention.

Referring to FIG. 6, quad tree partitioning may be described for a scene change candidate block selected in the previous frame 10 and the current frame 20. FIG.

Specifically, when the scene change candidate block is divided into quad trees, it is possible to divide the scene change candidate block into four sub-blocks again. As in the step of determining the scene change candidate block in the previous sub-block, The scene change candidate block may be selected by comparing it with the threshold value and / or the second threshold value.

Referring again to FIG. 1, after determining a scene change candidate block (S120), the determined scene change candidate block is divided into quad trees, and each divided block is compared with a block of a previous frame corresponding thereto And selecting the scene change candidate block.

In addition, if it is possible to generate the interpolation frame by motion estimation and compensation as in the motion compensation frame rate enhancement technique considering motion, it may not be necessary to detect the scene change. Therefore, in such a case, it may be advantageous to judge whether or not to exclude it from the scene change candidate block.

Accordingly, after the step of determining the scene change candidate block (S120), the step of reconstructing the scene change candidate block may exclude the block that can be predicted through motion estimation from the scene change candidate block.

Here, a method of finding a block most similar to a scene change candidate block using motion estimation can be determined as follows.

Referring to Equation (10), a block is compared between the current frame f _cur (x, y) and another reference frame f _ref (x + dx, y + dy) to minimize the sum of absolute differences The most similar block can be determined.

7 is an exemplary diagram for explaining directional distribution analysis according to an embodiment of the present invention.

Referring to FIG. 7, a process of finally determining a scene change block among scene change candidate blocks through directional distribution analysis can be described.

The directional distribution analysis may be a process of determining a final scene change block based on scene change ratios of surrounding blocks among the found scene change candidate blocks.

Specifically, for each of the upper left blocks 30, the upper left blocks 31, the lower left blocks 32, and the lower right blocks 33 with the scene change candidate block CBi _{, j} as a center, It is possible to judge whether or not to switch. Herein, each block may have a value of 0 or 1 indicating whether or not a scene change is to be made.

This can be expressed as the following equation.

), 우상단블록들에 대한 평균(

), 좌하단블록들에 대한 평균(

), 우하단블록들에 대한 평균(

)을 도출할 수 있다.Referring to Equation (11), it is possible to derive the results of calculating the average of the scene transition states of the neighboring blocks of the scene change candidate block. The average values of the upper left blocks

), The average for upper-right blocks

), The average of the lower left blocks (

), The average for the lower right blocks (

Can be derived.

The scene change block can be finally selected as follows using the values thus derived.

Referring to Equation (12), if any one of the averages of the neighboring blocks derived above is larger than the third threshold value T3, it can be determined as a scene change block (FB _{i, j} ) If not, it can be set to 0.

8 is a block diagram of an adaptive scene change detection apparatus according to an embodiment of the present invention.

8, the adaptive scene change detection apparatus 40 may include a processor 41 for executing at least one instruction and a memory 42 for storing at least one instruction. have.

The adaptive scene change detection device 40 may include a storage 43 for storing video data for detecting a scene change and may include a communication module 44 for receiving video data or transmitting a scene change detection result .

Here, the adaptive scene change detection device 40 may be, for example, a desktop computer, a laptop computer, a notebook, a smart phone, a tablet PC, A mobile phone, a smart watch, a smart glass, an e-book reader, a portable multimedia player (PMP), a portable game machine, a navigation device, a digital camera, Digital multimedia broadcasting (DMB) players, digital audio recorders, digital audio players, digital video recorders, digital video players, personal digital assistants (PDAs) And so on.

Here, the processor 41 extracts a luminance component in each of the current frame and the previous frame of the current frame, performs binary coding on the luminance component based on the predetermined block according to the resolution in the current frame and the previous frame, The scene change candidate block of the current frame is determined by comparing corresponding blocks with respect to the luminance component of the current encoded frame and the previous frame, and the scene change block is selected by performing directional distribution analysis on the scene change candidate block.

Here, the processor 41 may filter the luminance components extracted from the current frame and the previous frame using a low-pass filter.

Here, the predetermined block according to the resolution may be determined to have a size proportional to the horizontal resolution of the current frame.

에 의해 결정된 블록 폭(BkSize)을 이용하여 BkSize× BkSize 의 크기를 갖도록 결정될 수 있다.Here, the predetermined block according to the resolution is expressed by the following equation

BkSize < / RTI >< RTI ID = 0.0 > BkSize < / RTI >

Here, the processor 41 may determine a scene change candidate block by comparing an average of pixel value differences between blocks of the current frame and blocks of the previous frame with a first threshold value.

Here, the processor 41 can XOR the pixel values of blocks between the current frame and the previous frame, and determine the scene change candidate block by comparing the average of the XOR values with the second threshold value.

Here, the first threshold value may be adaptively determined according to a size of a predetermined block.

Here, the first threshold value may increase linearly as the size of a predetermined block increases.

Here, the processor 41 may divide the determined scene change candidate block into quad trees and compare the divided blocks with blocks of the previous frame corresponding thereto, thereby selecting scene change candidate blocks.

Here, the processor 41 can reconstruct a scene change candidate block by excluding blocks predictable through motion estimation from the scene change candidate blocks.

The performance evaluation results using the adaptive scene change detection method or apparatus according to an embodiment of the present invention are as follows.

Table 2 compares PSNR according to scene change detection with conventional methods.

Referring to Table 2, [2] is a scene change detection method using a conventional multi-histogram, and [6] is a scene change detection method using a luminance component encoding. [4] is a technique using bidirectional motion estimation in frame rate enhancement. A, B, and C may denote the type of image, and CIF and Class C may denote the resolution (or format) of the image. The proposed scene transition detection method (Proposed) shows that the PSNR (in dB) is improved in the above Table 2 and the conventional methods.

Table 3 compares the execution time with the existing methods.

Referring to Table 3, it can be seen that the execution time (sec) is shorter than that of the conventional methods.

The methods according to the present invention can be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the computer readable medium may be those specially designed and constructed for the present invention or may be available to those skilled in the computer software.

Examples of computer-readable media include hardware devices that are specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions may include machine language code such as those produced by a compiler, as well as high-level language code that may be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate with at least one software module to perform the operations of the present invention, and vice versa.

Also, the above-described method or apparatus may be implemented by combining all or a part of the configuration or function, or may be implemented separately.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that

Claims (20)

Extracting a luminance component in each of a current frame and a previous frame of the current frame;
Performing binary coding on the luminance component based on a predetermined block according to a resolution in each of the current frame and the previous frame;
Determining a scene change candidate block of the current frame by comparing corresponding blocks with the luminance component of the current frame and the previous frame that are binary coded; And
And performing a directional distribution analysis on the scene change candidate block to select a scene change block. In claim 1,
After the step of extracting the luminance component,
Further comprising filtering each of the luminance components extracted from the current frame and the previous frame using a low pass filter. In claim 1,
Wherein the predetermined block according to the resolution comprises:
Wherein the frame is determined to have a magnitude proportional to the horizontal resolution of the current frame. In claim 1,
Wherein the predetermined block according to the resolution comprises:
Is determined to have the size of BkSize x BkSize using the block width (BkSize) determined by the following equation,

Wherein the Video_reso_width is the horizontal resolution of the current frame. In claim 1,
Wherein the step of determining the scene change candidate block comprises:
Determining an average of pixel value differences between a block of the current frame and a block of the previous frame with a first threshold value. In claim 1,
Wherein the step of determining the scene change candidate block comprises:
And XORing a pixel value between blocks of the current frame and blocks of the previous frame and comparing the average value of XOR (exclusive OR, exclusive-OR) with a second threshold value to determine an adaptive scene Transition detection method. In claim 5,
Wherein the first threshold value is a first threshold value,
Wherein the adaptive scene change detection method is adaptively determined according to the size of the predetermined block. In claim 7,
Wherein the first threshold value is a first threshold value,
Wherein the predetermined block size is linearly increased as the size of the predetermined block is increased. In claim 1,
After determining the scene change candidate block,
Further comprising the steps of: dividing the determined scene change candidate block into quad trees, and comparing each of the divided blocks with blocks of the previous frame corresponding thereto, thereby selecting scene change candidate blocks.
In claim 1,
After determining the scene change candidate block,
Further comprising reconstructing a scene change candidate block by excluding blocks predictable through motion estimation from among the scene change candidate blocks. A processor for executing at least one instruction; And
And a memory for storing the at least one command, the adaptive scene change detection apparatus comprising:
The processor comprising:
A luminance component extraction unit extracting a luminance component in each of a current frame and a previous frame of the current frame, performing binary coding on the luminance component based on a predetermined block according to a resolution in each of the current frame and the previous frame, Determining a scene change candidate block of the current frame by comparing the current frame and the previous frame with corresponding blocks of the predetermined block, performing a directional distribution analysis on the scene change candidate block to select a scene change block, Ever scene change detection device. In claim 11,
The processor comprising:
And filters each of the luminance components extracted from the current frame and the previous frame using a low-pass filter. In claim 11,
Wherein the predetermined block according to the resolution comprises:
Wherein the adaptive scene change detection unit is determined to have a magnitude proportional to the horizontal resolution of the current frame. In claim 11,
Wherein the predetermined block according to the resolution comprises:
Is determined to have the size of BkSize x BkSize using the block width (BkSize) determined by the following equation,

Wherein the Video_reso_width is the horizontal resolution of the current frame. In claim 11,
The processor comprising:
Wherein the scene change candidate block is determined by comparing an average of pixel value differences between a block of the current frame and a block of the previous frame with a first threshold value. In claim 11,
The processor comprising:
Wherein the scene change candidate block is determined by performing an XOR operation on a pixel value between blocks of the current frame and blocks of the previous frame and comparing an average value of XOR (exclusive OR, exclusive-OR) Ever scene change detection device. In claim 15,
Wherein the first threshold value is a first threshold value,
Wherein the adaptive scene change detection unit is adaptively determined according to the size of the predetermined block. In claim 17,
Wherein the first threshold value is a first threshold value,
Wherein the block size is linearly increased as the size of the predetermined block is increased. In claim 11,
The processor comprising:
Wherein the scene change candidate block is divided into quad trees, and each of the divided blocks is compared with a block of the previous frame corresponding thereto to select a scene change candidate block. In claim 11,
The processor comprising:
Wherein the scene change candidate block is reconstructed by excluding blocks predictable through motion estimation from the scene change candidate blocks.

Images