KR101526490B1 - Visual data processing apparatus and method for Efficient resource management in Cloud Computing - Google Patents

Abstract

The present invention relates to a visual data processing apparatus and method for efficiently utilizing resources of a thin client in a cloud computing system, and more particularly, , It is aimed to reduce the redundant data and improve the processing speed by classifying and processing the screen area into the character and non-character block.

Description

Technical Field [0001] The present invention relates to a visual data processing apparatus and method for efficient resource management in cloud computing,

And more particularly, to a visual data processing apparatus and a method of operating the visual data processing apparatus for providing visual services in cloud computing.

Recently, the growth of mobile devices such as smartphones and tablet PCs is increasing exponentially and is expected to increase further. However, while mobile computing offers convenience and flexibility, there are limitations such as slow processors, low memory capacity, short battery life, and slow data transfer rates. In order to overcome the limitations of this limited mobile device, a variety of mobile cloud services are emerging that can support resources of cloud servers with infinite resources by applying thin-client technology based on cloud computing. In the initial mobile cloud, It started from using cloud resources such as storage by accessing cloud service through app, but now it is developing as a technology that utilizes cloud computing using mobile terminal.

The present invention effectively utilizes the resources of a limited mobile device through cloud computing to provide advanced computing devices such as a smart phone and a tablet PC with high computing power and fast transmission speed through cloud computing And also to utilize high-performance computing resources at anytime and anywhere in terms of the efficiency of the visual data and the resources of the mobile device.

The apparatus for processing visual data in cloud computing according to an embodiment of the present invention includes a change detection unit detecting a frame change between received time data, capturing a frame in which a change is detected, and generating image data; A format classification unit for dividing the image data generated by the change detection unit into a plurality of blocks and classifying the image data into a character block and a non-character block; And an encoding unit for compressing the character block and the non-character block classified by the format classification unit according to a predetermined compression method for each format.

According to an embodiment, the format classification unit may classify a plurality of blocks obtained by dividing the generated image data into a character block or a non-character block by using a boundary detection result that performs boundary detection in block units, Wherein the boundary detection unit divides the pixels of the block into boundary pixels and non-boundary pixels according to whether the boundary is detected, and the format classification unit divides the block into the character block or the non-boundary pixel by using the number of the boundary pixels and non- It can be classified as a non-character block.

According to an embodiment, the format classification unit may determine a character block when the ratio of the number of the boundary pixels to the total number of pixels is larger than a set value, and a non-character block when the ratio is smaller than a predetermined value, The presence or absence of the reference data is determined according to a predetermined algorithm. If there is no reference data, the boundary detection operation is performed on all the received blocks for all the blocks. If there is reference data, The neighboring blocks can be grouped based on the reference data to reduce the number of boundary detection operations.

A method of processing visual data in cloud computing according to an exemplary embodiment of the present invention includes: detecting a frame change of received time data; Capturing a frame in which the change is detected to generate image data;

Dividing the image data into a plurality of blocks; Classifying the block into a character block and a non-character block through a predetermined algorithm predetermined for the block; And compressing the classified character block and the non-character block by a predetermined compression method corresponding to each format.

According to an embodiment of the present invention, the step of classifying the image data into the character block and the non-character block may include a step of detecting a boundary of a plurality of blocks obtained by dividing the generated image data, Block, and the boundary detection result displays the pixels of the block as boundary pixels and non-boundary pixels according to the detection of the boundary,

The classifying into the character block and the non-character block may classify the block into the character block or the non-character block using the number of the boundary pixels and the non-boundary pixels, respectively.

According to an embodiment, the step of classifying into a character block and a non-character block may include determining whether a ratio of the number of the boundary pixels to the total number of pixels is greater than a set value, Wherein the step of classifying the character block into the character block and the non-character block includes determining whether or not the reference data is present according to a predetermined algorithm,

If there is no reference data, the boundary detection operation is performed on all the received blocks with respect to the plurality of received blocks. If there is reference data, neighboring blocks are grouped based on the reference data for the received plurality of blocks The number of boundary detection operations can be reduced.

1 is a diagram showing a relationship between a visual data processing device and a peripheral device according to an embodiment.
2 is a configuration diagram of a visual data processing apparatus according to an embodiment.
3 is a block diagram of the format classification unit of FIG. 2 according to an embodiment.
4 is a block diagram of the boundary detection unit of FIG. 3 according to an embodiment.
5 is an algorithm for a format classification method using a Sobel technique used in the format classification unit of FIG. 3 according to an embodiment.
6 is a flowchart of a visual data processing method according to an embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

Throughout the specification, when an element is referred to as "comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise.

Hereinafter, a data processing apparatus and method for efficient data transmission in cloud computing according to an embodiment of the present invention will be described with reference to the drawings.

1 is a diagram showing a relationship between a visual data processing apparatus 1000 and a peripheral device according to an embodiment.

Referring to FIG. 1, the visual data processing apparatus 1000 may perform operations of data from the data providing apparatus 2000, output the results of the operations, and receive the generated visual data.

The generated time data can be used without limitation of the format. However, when the function of the visual data processing apparatus 1000 to be described later is taken into consideration, the calculated result value is a mixture of images (image data) and characters The visual data having the format can be used mainly because the utilization efficiency is the highest.

One embodiment of using time data having a format in which images and characters are mixed is time data for an Internet web page. Web pages can be performed on mobile devices without using cloud computing. However, cloud computing can be used to provide information to users in order to efficiently use limited resources such as computing power and battery capacity.

According to one embodiment in which the data of the web page is used as the time data to be processed by the time data processing apparatus 1000, the time data processing apparatus 1000 compares the time data which is an output result value received from the data providing apparatus 2000 Thus, the existing frame and the unchanged frame can use the previous frame without transmission, and can capture the frame when the changed frame is detected.

Thereafter, the captured image data is divided into a plurality of blocks and selected as non-character blocks and character blocks.

After encoding by applying different compression methods according to the selected format, it is transmitted to the data output apparatus 3000.

 Herein, the data output apparatus 3000 mainly uses a thin client, and the thin client can be a mobile device such as a smart phone, a PDA, a tablet PC, and the like.

The reason why the data providing apparatus 2000 performs calculation of data and transmits the resultant value to the visual data processing apparatus 1000 is as follows.

 In general, since the time data has a large capacity and it requires a high data operation capability in calculating and outputting the data, mobile devices such as a smart phone, a PDA, and a tablet PC, which are mainly used as a thin client, In processing the data, problems such as a data storage space, a lack of computing power, a delay in operation speed, and an excessive consumption of the battery may occur.

In order to solve such a problem, a virtual machine (VM) or a thick client of a server in which there is no problem such as a data storage space, an arithmetic capability, a battery, etc., The time information can be transmitted to the user even in an environment having a small storage capacity and low computation capability.

2 is a configuration diagram of a visual data processing apparatus 1000 according to an embodiment.

2, the visual data processing apparatus 1000 may include a change detection unit 1100, a type classification unit 1200, and an encoding unit 1300,

The change detecting unit 1100 receives the time data generated by the data providing apparatus 2000 as the result of performing the calculation of the time data and determines whether the received time data is the same as the previously received frame Or whether there is a change.

When the change is not detected, it is possible to transmit the duplicated data without passing the data to the data output apparatus 3000 because there is no reason to increase the data transmission capacity.

On the other hand, when a change in the frame is detected, the frame in which the change is detected can be captured to generate image data.

Here, the main purpose of the visual data processing apparatus 1000 is to receive time data from the data providing apparatus 2000, process the data, send the data to the data output apparatus 3000, It means that the received data is stored, processed and sent to the data output apparatus 3000 is not performed.

The reason for generating the image data is not to compress the entire time data received but divides the image data into a plurality of blocks into character and non-character blocks and compresses them, So that the load on the data processing of the data output device can be reduced and efficient resource utilization can be achieved.

The format classifying unit 1200 divides image data generated by capturing a frame in which the change generated by the change detecting unit 1100 is detected into a plurality of blocks and subjects each block to a character block or non- It is possible to judge whether or not it is a character block.

According to the discrimination result, the block type can be classified into a character block or a non-character block.

Hereinafter, a predetermined algorithm for determining a block and classifying a format will be described in detail with reference to FIG.

The encoding unit 1300 may compress the format of the block into a character block or a non-character block in the format classification unit 1200, compress the format of the block according to each format, and output the compressed data to the data output apparatus 3000 .

Here, the compression method suited to each format is not limited to any particular compression method, but can be used without limitation as long as it is an efficient compression method in a given environment.

FIG. 3 is a configuration diagram of the type classification unit 1200 of FIG. 2 according to an embodiment.

Referring to FIG. 3, a data divider 1210, a boundary detector 1220, and a selector 1230 may be included.

The data dividing unit 1210 may divide the image data generated by capturing the frame detected by the change detecting unit 1100 into a plurality of blocks of the same size.

In this case, the size and the number of blocks may vary according to the user's setting value. As the size of the block increases, the number of blocks decreases. As the number of blocks decreases, the time required to detect a boundary is reduced. The accuracy of non-character region search results is reduced.

On the other hand, as the size of the block decreases, the number of blocks increases. As the number of blocks increases, the time required to detect the boundary increases, and the accuracy of the character and non-character region search results increases.

Here, the reason why the relation between the number of blocks and the time required for the detection of the next boundary is in the proportional relationship is that the detection operation must be performed for each block through a predetermined boundary detection algorithm in order to detect the boundary. Time is also increased.

The reason why the relationship between the number of blocks and the accuracy of the character and non-character area search results is in proportion is that the size of the block decreases as the number of blocks increases and the area as smaller as the size of the block decreases Since a detection operation can be performed, more accurate detection becomes possible.

The boundary detecting unit 1220 may use a method of detecting a boundary through a predetermined algorithm to discriminate a plurality of blocks generated by the data dividing unit 1210 into a character and a non-character block.

The reason for using the method of detecting the article boundary is that since the character has mostly the vertical and horizontal boundaries as attributes, it is possible to detect the boundary and recognize the character area through the detection result.

Here, the predetermined algorithm means various techniques for detecting the boundary, and in one embodiment, the boundary can be detected using the Sobel technique.

The Sobel technique, also called Sobel Edge Detection, is used to detect the edge of an image, ie, the boundary. After calculating the slope of the vertical direction and the slope of the horizontal direction, the slope of the result calculated through the Sobel technique can be expressed as Equation (1).

[Equation 1]

Where A represents the image to be processed, and Gx and Gy represent the slopes in the vertical and horizontal directions, respectively. Using Gx and Gy, G, the size of the boundary, can be obtained as shown in Equation (2).

&Quot; (2) "

However, in the actual implementation, since the calculation using Equation 2 is computationally complicated and takes a long time, the method using absolute value such as Equation 3 is mainly used.

&Quot; (3) "

The boundaries of image data can be detected using this Sobel technique. At this time, the portion detected as a boundary is divided into a non-boundary pixel by a boundary pixel and a non-boundary pixel.

The selection unit 1230 may select each block as a character block or a non-character block through a predetermined algorithm with a result of detecting the boundary in the boundary detection unit 1220. [ Here, the predetermined algorithm will be described in detail with reference to FIG.

FIG. 4 is a block diagram of the boundary detector 1220 of FIG. 3 according to an embodiment.

4, the boundary detecting unit 1220 may include a reference data checking unit 1221, an initial detecting unit 1222, and a continuation detecting unit 1223. [

The reference data checking unit 1221 can determine whether there is data to be used as a reference for classification of the type of the block with respect to the data received by the boundary detecting unit, whether it is a character block or a non-character block.

Here, the reference data refers to data that includes format classification results of a block already obtained through the boundary detection operation and the type classification operation in a specific section in order to classify the format of a plurality of divided blocks .

Through the boundary detection result, it is not necessary to perform a boundary detection operation for all the blocks, and it is possible to perform less boundary detection by grouping the neighboring blocks of the block determined as the boundary.

In order to determine whether or not reference data exists, a method is used in which a section is specified through a certain reference among continuous data, and reference data corresponding to the section exists.

Various methods can be used to select the segmentation criterion. In the data connection protocol, a frame synchronization scheme is used in which a special bit combination is used to indicate the start and end of a frame, and a synchronization is established between the transceivers in this manner But is not limited to.

If there is no reference data in a specific section, the initial detection section 1222 can perform a boundary detection operation using a boundary detection method such as Sobel technique for all of a plurality of blocks obtained by dividing the generated image data.

The reason for performing the boundary detection operation on all the blocks is that the boundary detection operation is performed on all the blocks because there is no reliable data that can be used as a reference since the change is detected in the first frame.

If the reference data exists in a specific section, the consecutive detecting section 1223 detects a boundary detection method such as the Sobel method on a smaller number of blocks than the first detecting section 1222 among a plurality of blocks obtained by dividing the continuous image data The boundary detection operation can be performed.

Here, since the consecutive detection unit 1223 has reliable data that can be used as a criterion through the boundary detection operation for a specific section, the sorting unit 1230 selects the character or non- It is assumed that neighboring blocks of the selected block have a similar block shape, and the neighboring blocks are grouped according to the existing selection results instead of all of the blocks to perform less boundary detection operation .

According to an embodiment, the number of detection operations can be reduced to 1/4 by grouping neighboring blocks of a block identified as a character, thereby reducing a computation speed and a load on an operation.

Herein, 1/4 is only one embodiment, and it is not limited to this, as long as it can reduce the number of detection operations and have an accuracy with respect to a constant detection result.

FIG. 5 is an algorithm for format classification using the Sobel technique used in the format classifier 1200 of FIG. 3 according to an embodiment.

Referring to FIG. 5, an algorithm for a data classification method using the Sobel technique is shown as follows.

The image data generated by the change detecting unit is divided into a plurality of blocks (S10).

Before performing the boundary detection operation, the user sets a setting value T, which is a criterion for discriminating the character or non-character block (S20).

Here, the set value T is a criterion for comparing the boundary detection result of the block with the value (I) obtained by dividing the block detection result into a character block when T is smaller and a non-character block when T is larger.

The reason for this comparison with the T value is that, in the case of a block having many vertical or horizontal boundaries, it is likely to be a character region.

After the set value T is established, the Sobel boundary detection operation is performed for each block (S30).

In operation S40, the region detected as a boundary is displayed as a boundary pixel, and the remaining region is displayed as a non-boundary pixel in operation S40. An I value obtained by dividing the number of boundary pixels by the block size is calculated and compared with the T value in operation S50.

If the I value is greater than the T value, it is determined to be a character block (S60). If the I value is smaller than the T value, it is determined to be a non-character block (S65).

When the determination process is completed, it is checked whether the target block remains, and if there is a remaining block, the boundary detection operation is repeated for each block (S70).

After the format of all of the target blocks belonging to the specific section is determined, the blocks of the same type are blocked (S80).

6 is a flowchart of a visual data processing method according to an embodiment.

Referring to FIG. 6, first, time data including a result of computing and outputting data is received through the data providing apparatus 2000, and a previous frame and a current frame are compared with each other, (S110).

If a change is not detected at this time, duplicate transmission of the same frame increases the total transmission capacity and requires a repeated associative operation, so that it can be passed without storing and transmitting data (S115).

If the previous frame is compared with the current frame and a change is detected, the current frame may be captured to generate image data (S120).

Here, the reason for generating the image data is to divide the image data into a character block or a non-character block for the purpose of increasing the compression ratio in order to reduce the transmission capacity and the computational effort.

The image data thus generated can be divided into a plurality of blocks (S130).

In this case, the size of the block and the number of blocks may be determined according to the setting of the user. In one embodiment, the blocks may be divided into 16x16 blocks so that the images do not overlap. Unless the image data is divided into a plurality of blocks and the image data is excessively divided so that the boundary detection operation takes too much time or the image data is divided into a small amount and an accuracy is not reduced, It can be divided into blocks.

In operation S140, the boundary detector 1220 receives a plurality of blocks thus produced, and determines whether there is reference data for a specific section including a plurality of received blocks.

If there is no reference data, the boundary detection operation can be performed on all the plurality of generated blocks (S150).

Since this is the case of the first reception, there is no dictionary information on the positions of the character and non-character blocks, so that the boundary detection operation is performed on all the blocks.

The boundary detection operation is performed because the boundary detection operation can search for characters having vertical and horizontal boundaries as attributes and divide them into character blocks and non-character blocks.

The boundary detection method may use the Sobel method according to an embodiment of the present invention. However, the present invention is not limited thereto, and the method can be used when there is a more efficient method for edge detection.

On the other hand, when the reference data exists, since there is reliable data that can be used as a reference through the boundary detection operation, neighboring blocks of the selected block based on the classification result of the existing data have almost the same block form In step S155, the neighboring blocks are grouped according to the existing selection results, not all of the blocks, so that less boundary detection is performed.

According to an embodiment, the number of detection operations can be reduced to 1/4 by grouping the neighboring blocks of the blocks identified as characters, thereby reducing the computation speed and the load on the computation. But the present invention is not limited thereto as long as it can reduce the number of detection operations and have an accuracy with respect to a constant detection result.

With the result of the boundary detection thus derived, a plurality of blocks are classified into a character block or a non-character block according to a predetermined algorithm shown in FIG. 5 (S160).

The blocks classified into the character or non-character block are compressed using a compression method suitable for each format (S170).

The compression method used here does not mean a specific method, but can be used without limitation as long as it is capable of achieving a high compression ratio and compression quality in compressing data of a character or non-character format.

The compressed character or noncharacter data may be divided into a character or noncharacter area that is not a block unit, and may be transmitted to the data output apparatus 3000 through the entire area and output to the user.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Various modifications and improvements of those skilled in the art using the basic concept of the present invention are also within the scope of the present invention.

1000: Visual data processing device 2000: Data providing device
3000: Data output device 1100: Change detector
1200: type classification unit 1300: encoding unit
1210: Data partitioning unit 1220:
1230: Selection unit 1221: Reference data verification unit
1222: initial detection unit 1223:

Claims (10)

A change detecting unit detecting a frame change between received time data, capturing a frame in which a change is detected, and generating image data;
A format classification unit for dividing the image data generated by the change detection unit into a plurality of blocks and classifying the image data into a character block and a non-character block; And
And an encoding unit for compressing the character block and the non-character block classified by the format classification unit according to a predetermined compression method for each format,
Wherein the type classification unit classifies a plurality of blocks obtained by dividing the generated image data into a character block or a non-character block by using a boundary detection result obtained by performing boundary detection on a block-by-block basis. Device. delete The method according to claim 1,
The boundary detection result displays the pixels of the block as boundary pixels and non-boundary pixels according to whether the boundary is detected,
Wherein the type classification unit classifies the block into the character block or the non-character block by using the number of the boundary pixels and the non-boundary pixels, respectively. The method of claim 3,
Wherein the format classification unit determines the character block as a non-character block if the ratio of the number of the boundary pixels to the total number of pixels is larger than the set value, and if the ratio is smaller than the set value. 2. The apparatus of claim 1, wherein the format classification unit determines the presence or absence of reference data according to a predetermined algorithm,
If there is no reference data, the boundary detection operation is performed on all the received blocks with respect to the plurality of received blocks. If there is reference data, neighboring blocks are grouped based on the reference data for the received plurality of blocks And the number of boundary detection operations is reduced. Detecting a frame change of the received time data;
Capturing a frame in which the change is detected to generate image data;
Dividing the image data into a plurality of blocks;
Classifying the block into a character block and a non-character block through a predetermined algorithm predetermined for the block; And
And compressing the classified character block and the non-character block by a predetermined compression method corresponding to each format,
Wherein the classification into the character block and the non-character block is performed by classifying a plurality of blocks obtained by dividing the generated image data into a character block or a non-character block by using a boundary detection result obtained by performing boundary detection on a block- visual data processing method in cloud computing. delete The method according to claim 6,
The boundary detection result displays the pixels of the block as boundary pixels and non-boundary pixels according to whether the boundary is detected,
Wherein the classification into the character block and the non-character block classifies the block into the character block or the non-character block using the number of the boundary pixels and the non-boundary pixels, respectively. 9. The method of claim 8, wherein classifying into the character block and non-
Determining whether the ratio of the number of the boundary pixels to the total number of pixels is larger than the set value, and determines the character block as a non-character block if the ratio is smaller than the set value; The method according to claim 6, wherein the step of classifying the character block into non-character blocks comprises determining whether there is standard data according to a predetermined algorithm,
If there is no reference data, the boundary detection operation is performed on all the received blocks with respect to the plurality of received blocks. If there is reference data, neighboring blocks are grouped based on the reference data for the received plurality of blocks A method for processing visual data in cloud computing by reducing the number of boundary detection operations.

Images