WO2004039081A1

WO2004039081A1 - Real time lossless compression and restoration method of multi-media data and system thereof

Info

Publication number: WO2004039081A1
Application number: PCT/KR2002/001983
Authority: WO
Inventors: Kil Soon Park
Original assignee: Boram C& C Co., Ltd
Priority date: 2002-10-24
Filing date: 2002-10-24
Publication date: 2004-05-06
Also published as: AU2002354510A1

Abstract

The present invention relates generally to a method for lossless data compressing and decompressing. Specifically, this invention relates to a system and a method for real-time lossless multimedia data compressing and decompressing for transmitting data over Internet having an unstable bandwidth. A moving picture data is transformed into a joint photographic experts group (JPEG) code via a hardware and a voice/video synthesis rate is controlled based upon a client's network data transmit capacity. A certain amount of data bit stream is classified with the bit unit and a bit group is formed. The data is compressed via replacing each bit group with “1” or '0' and the compressed data is decompressed via inverse-replacing. A certain amount of data bit stream is divided in a certain number of bits. A bit group is formed, each group is replaced with '1' ('1' or '0') and is decompressed into original status of the data.

Description

REAL TIME LOSSLESS COMPRESSION AND RESTORATION METHOD OF

MULTI-MEDIA DATA AND SYSTEM THEREOF

TECHNICAL FIELD The present invention relates generally to a method for lossless data compressing and decompressing. Specifically, this invention relates to a system and a method for real-time lossless multimedia data compressing and decompressing for transmitting data over Internet having an unstable bandwidth. A moving picture data is transformed into a joint photographic experts group (JPEG) code via a hardware and a voice/video synthesis rate is controlled based upon a client's network data transmit capacity. A certain amount of data bit stream is classified with the bit unit and a bit group is formed. The data is compressed via replacing each bit group with "1" or "0" and the compressed data is decompressed via inverse-replacing.

BACKGROUND ART

Recently a network such as local area network (LAN) or wide area network (WAN) is popularized and a communication skill and a data compressing skill are rapidly developed. Therefore, a multimedia service, which allows users to feel diverse information such as pictures, graphics and sounds etc. simultaneously and correctly, is activated and realized. This multimedia service is used for various fields such as a real-time Internet broadcasting service, a video on demand

(VOD) , a videoconference and an internal broadcasting station.

The size of multimedia data including a moving picture file is big so the data should be compressed, transmitted, decompressed and played via a multimedia player for a real-time monitoring or a broadcasting over network.

However, there is a limitation to improve a data transmit speed via increasing a compress rate because the quality of moving picture file degrades with the compress rate.

The official regulation of "data transmission format"

(data creating - data transmitting - data playing) and the definition of "time" have not been existed in a real-time moving picture file transmitting system over Internet until now. So an acceptable time delay for the data creating, data transmitting and data playing in the case of Internet videoconference should be within 1 second for a client . In this case, each time for data creating and data transmitting should be within 0.5 second so an uncreated data within 0.5 second should be eliminated for the next data transmitting. Therefore, a data loss is unavoidable in the real-time videoconference over

Internet having an unstable bandwidth. There is no way to prepare an enough bandwidth for a lossless data transmitting because either data loss or data traffic via unexpected causes is an unavoidable Internet characteristics.

Therefore, an appropriate data format is necessary in Internet environment . In order to transmit a data over Internet , the data is sequentially transmitted via the unit of bit and is consisted of a head field, a data field, and a trailer field in succession. The header is consisted of a flag, an address field and a data field, and the trailer field is consisted of frame certificate number flags. Information or stored data in the frame may include a capsule type frame that is used in a higher layer or another protocol. Actually, "frame relay frame" generally transmits the framed data via previous protocol program. A "frame" in the field of "time division multiplexing" means an accomplished event cycle within a time division period. A "frame" in the fields of a film and a video recording/playing means just single image among series of images that are recording and playing. A "frame" in the field of computer video playing technique means an image which is transmitted to a video image rendering system and the image is modified and played consequently in a frame buffer of video memory .

Internet users have been troubled with understanding the contents of moving picture file, which is consisted of frames and is transmitted over Internet, due to the Internet unavoidable data loss problem. As an exchanging information such as a multimedia transmitting, a real-time broadcasting, a videoconference and a management of e-business has been accomplished and users' demand for high quality multimedia information has been highly desired so the amount of data has been enormously increased. Therefore, to improve a data transmit speed and a data accuracy, improvement of a computer data processing speed, maximization capacity of communication cable, a diversification of data compress techniques have been studied and developed. However, a lossy compress technique such as a motion picture export group (MPEG) or a joint photographic experts group (JPEG) technique is currently used so a high quality video or sound information is unavailable and a compress rate is drastically decreased with a data type. In addition, a conventional compress technique does not provide a recompressing process so there is a limitation to improve the compress rate. Specifically, as exchanging information in cordless communication is popularized, a high data compress rate and an exact data reproduction are highly demanded to decrease transmission load effectively.

DISCLOSURE OF THE INVENTION

This invention provides a lossless, high compressible, data independent and re-compressible compress method to a client for maintaining a high quality video information and for minimizing a time delay in data transformation, a data transmission and a data reproduction process in the network environment via a bit-replacing. The network environment has an unstable bandwidth as an Internet .

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 depicts a block diagram of transmitting procedure in the real-time lossless multimedia data compressing and decompressing system.

PIG. 2 depicts a block diagram of receiving procedure in the real-time lossless multimedia data compressing and decompressing system.

FIG. 3 depicts a transmitting procedure for the data of the real-time lossless multimedia data compressing and decompressing system.

FIG. 4 depicts a receiving procedure for the data of the real-time lossless multimedia data compressing and decompressing system. FIG. 5 depicts a flow diagram of a lossless data compressing process.

FIG. 6 depicts a flow diagram of a lossless data decompressing process From FIG. 7a to FIG. 7i depict an exemplary data table for the real-time lossless data compressing method.

From FIG. 8a to FIG. 8f depict an exemplary data table for the real-time lossless data compressing method.

BEST MODE FOR CARRYING OUT THE INVENTION

In accordance with this invention, a method and system for real-time multimedia data compressing and decompressing is described in below. FIG. 1 and FIG 2 depict a block diagram of a transmitting procedure and a receiving procedure in the real-time lossless multimedia data compressing and decompressing system, respectively.

A transmitting means 100 is consisted of a data storing part 110, a data memory part 120, a data dividing and sorting part 130, and a data transmitting part 140 as shown in FIG. 1.

The data storing part 110 stores voice data 3 and video data

4. The data memory part 120 having several buffers (121a, 121v) to store voice data 3 and video data 4 temporary and separately from a multiplex data (voice data 3 and video data 4) . The data dividing and sorting part 130 randomly sorts the stored data in the data memory part 120 for client 1 to discern the data.

The data transmitting part_. 140 having several sockets 141 transmits the randomly sorted data to the data receiving part 200 .

A receiving means 200 is consisted of a data receiving part 210, a data dividing part 220, a data memory part 230, a data sorting part 230 and a data extracting part 240. Data receiving part 210 receives the multiplex data (voice data and video data) which is transmitted from the data transmitting means 100. The data dividing part 220 divides the received multiplex data into voice data 3 and video data 4. The data memory part 230 has several buffers (231a, 231v) , that store the data temporary, to sort the divided data in a sequence. The data sorting part 230 sorts the multiplex data that is stored in the data memory part 230. The data extracting part 240 extracts a voice and an image from the sorted voice and video data for client 1 to hear and see. FIG. 3 and FIG. 4 depict a transmitting procedure and a receiving procedure for the data of the real-time lossless multimedia data compressing and decompressing system, respectively.

As described in FIG. 3 or FIG. 4, the real-time lossless compressing and decompressing system is consisted of a data extracting step SI and a data receiving step S5. The system divides the multiplex data, that is stored in the data storing part 110, into a voice data 3 and a video data 4, stores the divided data temporary in the data memory part 120, sorts the stored data in a random sequence via the data sorting part 130, and extracts the sorted data over Internet 2 via the data transmitting part 140 for a client 1 to discern the data contents at the data extraction step SI. The system receives the extracted multiplex data over Internet 2, divides the received data into voice data 3 and video data 4 via the data dividing part 220, stores the divided data in the data memory part 230 to sort the data in a sequence, and arrays the stored voice data 3 and video data 4 in a sequence, and extracts the arrayed data from the extracting part 240 for client 1 to hear and see the data even if some part of data is missing in the data receiving step S5.

A common memory 110 and buffers (121, 122) of data memory part 120 are initialized, a connecting status is confirmed via client 1, voice and image are captured at the initial step, a system initialization step S2 is executed to transmit a real-time moving picture over socket 141.

The voice data 3 and video data 4, which are supplied via the system initialization step S2 , are divided and the voice data 3 is captured, divided and stored for a priority transmission .

A frame is divided to transmit video data 4 in the rate of 30 frames per second, the divided frame is numbered, the data in the divided frame is selected and the corresponding number of the selected frame is stored for a priority transmission.

The frame is prepared to transmit the header, the voice data 3 , the priority transmit video data 4 and the remained video data 4 in a sequence via combining the voice data 4 and the video data 4 at a data creating step S3.

A data-transmitting step S4 transmits the combined voice data 3 and video data 4 to a waiting connected client 1.

Next steps are consisted of a data receiving step S5, a system initialization step Sβ, a data receiving and storing step S7, a voice data storing step S8 and a voice and video data decompressing step S9.

A Buffer 230 and a socket 141 are initialized, an extracting server is connected and a thread is created at the system initialization step S6. The multiplex data (voice data 3 and video data 4) is received via the data receiving part 210 and is stored in the diverse data memories 230. The type of the received data is decided between voice data or not, the initial value of the voice data is set up such as BF0=F, BF1=F and BF2=T, the predefined bit is set up with the initialized value, the voice is classified with each address via set flag and the serial number of video data is extracted. The extracted serial number is sorted at a data receiving and storing step

S7, the voice data 3 of S7 step is stored in memory 3 at the voice storing step S8, both the stored voice data 3 at each address and the stored video data 4 in serial number's order are combined and decompressed at the voice and video data decompressing step S9.

The data receiving and storing step S7 is consisted of 7 steps. The first step S7a receives and stores the multiplex data 3,4 after the system initialization step, the second step

S7b judges whether the multiplex data 3,4 is stored or not in the first step S7a, the third step S7c judges whether the multiplex data is voice data 3 or not, the fourth step S7d classifies the voice with initial value if the multiplex data is voice data 3 , the fifth step S7e appoints the classified voice data to the pr-defined bit/bits, the sixth step S7f classifies the appointed voice data 3 by the unit of bit and temporary stores the classified voice data 3 into buffer 231v, the seventh step S7g extracts the video data 4 after the third step S7c in a sequence, the eighth step S7h stores sequentially the extracted video data 4 into buffer 231v based upon the initial values of them.

FIG. 5 and FIG. 6 depicts flow diagrams of a lossless data compressing process and data decompressing process, respectively.

As described in FIG.5, the data compressing process using the bit substitution in accordance with this invention as follows: At first, an original data stream having n bits data amount is obtained.

And then the original data stream is classified in a k bit unit, (2 k n) , a target of compress data stream is consisted of bit groups . The bit group is consisted of k bits having one of values among 0 to 2^k-l and is classified with its value and the frequency of the bit group is checked with a class. For example, if k is 2 then a, b, c, d bit groups are classified as "10", "01", "11" and "00", respectively. A bit group having the highest frequency in a data stream is replaced with "1" and the other bit groups are replaced with "0" so the first compress table is formed.

After removing the bit group that is replaced with "1" , a bit group having the highest frequency in the data stream is replaced with "1" and the other bit groups are replaced with "0" so the second compress table is formed.

Above mentioned removing bit group process or replacing bit group process is applied to all bit groups and maximum 2^k-l compress tables are formed. The last 2 -bit groups can be formed as a compress table.

A final compress table can be formed with combining of all compress tables in the forming sequence. And a high compress rate can be accomplished via repeating above the mentioned compressing process. In accordance with this invention, the data compressing process is independent on the data type but dependent on the data value or the distribution of bit group so the data compressing process for compressed data is possible.

The data stream, which can not meet a minimum requirement for the compressing process, is deformed to improve the data compress rate because the data compress rate is dependent on the data value.

The compressed data is transmitted in the form of transmit bit frame as follows:

The compress information bit stream includes the size and array sequence of compress table and the size and substitution information of bit group, and the control bit stream includes the data deformation information for improving the data compress rate. The transmitting bit frame can be performed in diverse format on demands .

Meanwhile the lossless decompressing process (or compressing process) of compressed data is as follows: The compressed information and the compressed data stream is acquired through a transmitting bit frame as described in FIG. 4. The compressed data stream is divided into each compressed table with the size of compressed table and the array sequence information. The frame of decompressed table is formed from the first compressed table according to size of bit group. If the size of bit group is 2 -bits (k=2) , the decompress table has 2 times the number of bits of the first compress table.

According to the array sequence (the compressing sequence) and the substitution information, the decompress table is re-replaced with each compress table. If the compressed data is compressed several times, the compressed data is decompressed via decompressing process using the corresponding compress information. If a data is deformed, the data is re-transformed into an original condition via the control information. Accordance with this invention, a high compress rate and a data stream without data loss can be effectively acquired.

An execution example of the compressing/decompressing method is explained via data tables using a bit group having 2 -bits unit.

From FIG. 7a to FIG. 7i depict an exemplary data table for the real-time lossless data compressing method. As described in FIG 7a, an original data stream having a certain amount (20Obits) of data is extracted from video or voice information bit stream, etc. The original data stream is divided in 2 -bits unit and the first target of compressing data stream is formed as a total 100 bit groups as described in FIG 7b.

Frequency is calculated after recognition and classification of data contents. FIG 7b shows 50 a bit groups having "00", 25 b bit groups having "01", 20 c bit groups having "10" and 5 d bit groups having "il" . The a bit group having the highest frequency is replaced with "1" other b, c, d groups are replaced with "0" from the first target of compress data stream and then the first compress data table, j table, is created . The second target of compress data stream is recreated via removing a bit group in the first target of compress data stream as described in FIG 3d. The second compress data table, k table, is formed via replacing b group with "1" and replacing c, d groups with "0" from the second target of compress data stream as described in FIG 7e. The third target of compress data stream is formed via removing the replaced b group in the second target of compress data stream as described in FIG 7f . The third compress data table, 1 table, is formed via replacing c group with "1" and via replacing d group with "0" from the third target of compress data stream as described in FIG 7g. The compress data stream having 175 bits information is obtained via sequentially combining of each compress table as described in FIG 7h. Finally, 12.5% compressed data stream is acquired in the first example.

The compressed data stream is divided in 2 -bits unit and can be compressed again so high compress data rate can be obtained via performing several compress steps on demands as described in FIG 7i. From FIG. 8a to FIG. 8f depict an exemplary data table for the real-time lossless data compressing method using 2 -bit groups. The compress data stream is acquired from transmit bit frame as shown in FIG 8a and the compress data stream is classified as j , k, and 1 tables in a sequence with the compress information.

Decompress data table is created via acquiring 2 times number of bits in j table with substitution in 2 -bits unit and "1" in j table is replaced with "00" at corresponding position in decompress table. "1" in k table is sequentially inverse replaced with "01" in decompress table as described in FIG. 8d.

Decompress table is finally accomplished as described in FIG. 8e via replacing "1" with "10" and "0" with "11" in 1 table, respectively . Therefore, the 200 bits data stream, which is exactly same with original data stream, is decompressed as described in FIG 8f .

When a client 1 connects the data transmit server system, which transmits a real-time moving picture file for a video communication, a video conference and a real-time broadcasting etc., over Internet, the real-time data transmitting system initializes memory and frame buffer, captures voice and image and maintains their initial condition. And then the system divides voice data 3 and video data, stores the divided data into the data memory part 110, keeps the data in safe in case of telecommunication with computer in a distance.

If the data process capturing voice and video data is created, the stereo voice data 3 is temporary stored in several buffers (121a, 12lv) of the data memory part 120 and the divided voice data 3 and video data 4 is sorted via the data dividing and sorting part 130.

The voice data 3 is placed at the behind of frame header to be transmitted priority, and is divided for client 1 to hear the data contents at least while voice data 3 is missing over Internet 2 and stereo is impossible.

The video data 4 is divided into 30 frames per second for transmitting, the divided data is temporary stored in buffer (231a, 231v) of data memory part 230. For client 1 to discern the temporary stored data, a priority address is set up via the data dividing and sorting part 130, and the voice data 3 and video data 4 are transmitted over the data transmitting part 140.

The frame is randomly divided into 30 addresses, original number is assigned to the addresses, assigned numbers are randomly arrayed such as 1,16,9,24,5,20,

13,28,3,18,11, ,12,17,21,25,29 in a sequence and transmitted over Internet. While some part of the data is missing over Internet, for example ending part of data is missing, a client can understand the data contents because some initial, medium and ending part of data are still existed via the above sequence . The header, the voice data 3 and the video data 4 are positioned as a sequence in the frame and are extracted from the data transmitting part 140.

The multiplex data, voice/video data, is transmitted from the data transmitting part 100 to the data receiving part 210 of data receiving means 200.

The received multiplex data is transmitted to the data dividing part 220, is divided into voice data 3 and video data

4, the divided data are temporary stored in several buffers

(231a, 231v) of the data memory part 230 to sort sequentially. The temporary stored data is transmitted into the data sorting part 230, the video data 4 is sorted in the sequence of the randomly determined address and the voice data 3 is sorted with the video data 4. If the voice data 3 and video data 4 are sorted, the data are combined and extracted through speaker and monitor for client to hear and see.

This invention provides a useful data transmitting method for clients to understand data contents while some part of voice/video data is missing in Internet environment. The following is a detailed explanation through examples of the invention. It should be understood, however, that the detailed description and specific examples are given via way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

INDUSTRIAL APPLICABILITY

This invention compresses and decompresses data effectively without data loss and accomplishes high compress rate due to the data type independent and repeatable compressing process .

This invention allows a client to understand the data contents while video data is missing during video communication, videoconference and real-time broadcasting etc. over Internet having an unstable bandwidth.

Claims

1. A method for real-time lossless compressing/decompressing multimedia data comprising :

a data extracting step which further comprises the steps of: dividing a stored multiplexed data in a data storing part into a voice and a video data; storing temporary the divided data in a data memory part; sorting the temporary stored divided data in a random sequence via a data dividing and sorting part; and extracting the sorted data over Internet via a data transmitting part,

a data receiving step which further comprises the steps of: receiving the extracted data in a data receiving part; dividing the received data into a voice and a video data via a data dividing part ; storing the divided data in diverse data memory parts to sort the divided data; sorting the stored voice and video data in a memory part in a sequence; and extracting sorted data through an extracting part.

2. The method as claimed in claim 1, wherein said data extracting step comprising:

a system initializing step which further comprises the steps of: initializing both a common memory of a data storing part and a frame buffer of a data memory part; checking a connection status via a client; and capturing a voice and video data at an initial condition for real-time transmitting moving picture through a socket,

a data creating step which further comprises the steps of: dividing the received voice and video data from the system initializing step; capturing a voice data between the divided a voice data and a video data; dividing the captured voice data; storing the divided data for a priority transmitting; dividing a frame for a transmitting the captured video data, which was not captured at the capturing step, based upon 30 frames per second; assigning a number to the divided frame; sorting the data of divided frame for a client to understand the data contents; storing the corresponding number of the sorted data for a priority transmitting; combining the voice and the video data in order of a header, a voice data, a priority transmitting video data and other video data; and

a data transmitting step transmits the combined data to a client who is in a connection status.

3. The method as claimed in claim 1, wherein said data receiving step further comprising:

a system initializing step initializes a global buffer, a frame buffer and a socket, connects an extracting server, and creates a thread;

a data receiving and storing step which further comprises the steps of : receiving the extracted multiplexed data from a data extracting part via data receiving part; storing the received data in diverse data memories; determining an initial value of voice data such as BF0=F,

BF1=F and BF2=T etc. when the received data is a voice data; setting up pre-defined bit via the determined value; classifying a voice data via each address based upon a set flag; extracting a serial number of video data; sorting the extracted serial number in a sequence; arraying and verifying the serial number in a sequence;

a data storing step stores voice data of the data receiving and storing step into a memory; and

a decompressing step for the voice and video data comprises the steps of : combining the stored voice data in each address and the sorted video data in a serial number's order; decompressing the combined data.

4. The method as claimed in claim 3, wherein said data receiving and said data storing steps further comprising the steps of :

receiving the multiplex data from the system initializing step and storing the received data; checking whether multiplex data is received or not at the previous step; checking whether the checked data at the second step is voice data or not; sorting the voice data based upon the determined value at an initial stage when the checked data at the previous step is a voice data; setting up the sorted data via pre-defined bit/bits; classifying the set voice data in a unit of bit and temporary storing the sorted data in buffer; extracting the video data of the second checking step in the order of serial number; and storing the extracted video data into each buffer in the sequence of serial number based upon the determined value at the initial stage.

5. A method for real-time lossless compressing/decompressing multimedia data comprising the steps of :

receiving a certain amount of original data stream; classifying an original data stream into a bit group with a number of bit ; forming a target of compressing data stream; checking a frequency based upon the value of the bit group; forming a compress table via replacing the highest frequency bit group with a value having 1 bit value and via replacing the other bit groups with another value having 1 bit; forming a target of compress data stream via removing the highest frequency bit group from the table which is formed at the previous step; forming a compress table via the bit-replacing; forming another compress table via the bit-removing and bit-replacing; repeating the compress tables forming step until all compress tables are formed; and forming a compress data stream via combining the pre-formed compress data tables in a sequence.

6. The method as claimed in claim 5 , wherein said bit group is consisted of a 2 -bit unit, the replacing target bit group is replaced with "1" , the other bit groups are replaced with "0", and finally a compress data table is formed.

7.The method as claimed in claim 5 or 6, wherein both said removing process and said replacing process are applied to all bit groups in the target of compress data stream, and compress data tables of 2^k-l number at maximum are formed.

8. The method as claimed in claim 5, wherein said forming a target of compressing data stream step further includes a data deforming step when the condition of data stream does not meet with a minimum requirement for compressing.

9. The method as claimed in claim 5, wherein a bit stream containing said compressing information and a transmitting bit frame containing said compress data stream are formed.

10. The method as claimed in claim 5 or 9, wherein said compress data is consisted of the bit frame which is sorted an initial bit stream, a compress information bit stream, a control bit stream, a compress data bit stream and an end bit stream in a sequence .

11. The method as claimed in claim 10, wherein said compress information bit stream includes a size of compress table, a sorting sequence, a size of bit group and a replacing information.

12. The method as claimed in claim 10, wherein said control bit stream includes data deforming information for improving compress rate.

13. A method for real-time lossless compressing/decompressing multimedia data comprising the steps of :

receiving a compress information and a compress data stream; dividing the compress data stream into each compress table based upon the compress information; forming a format of data decompressing table based upon the first compress table; and reconstructing original data stream via inverse-replacing each compress table with reconstruction table in a sequence based upon the replacing information of the compress information.

14. The method as claimed in claim 13 , wherein the content of said bit group is recognized, the recognized data is classified with a category and each frequency of the classified data is calculated.

15. A system for real-time lossless compressing/decompressing multimedia data comprising:

a transmitting means which further comprises the steps of : dividing voice and video data from multiplex data for transmitting in Internet environment; extracting main part from the frame that is sorted via pre-assigned address; transmitting the extracted data preferentially; and transmitting the other data in a sequence,

a receiving means which further comprises the steps of : receiving the data that is transmitted via the transmitting means; dividing voice and video data; and combining and reconstructing the divided data in sequence .

16. The system as claimed in claim 15, wherein said transmitting means further comprising:

a data storing part which stores voice and video data; a data memory part which divides and temporary stores multiplex data that is transmitted through data storing part; a data dividing and sorting part which sorts the data, which is divided and stored temporary in data memory part, randomly; a data transmitting part which transmits the data that is divided and randomly sorted in data memory part into the receiving part .

17. The system as claimed in claim 15, wherein said receiving means further comprising:

a data receiving part which receives the multiplex data from the transmitting part; a data dividing part which divides the received multiplex data into both voice and video data; diverse memory parts which store the divided data temporary for sorting in a sequence; a data sorting part which sorts the data that is transmitted from the data memory parts; a data extracting part which extracts voice and video from sorted voice and video data for clients to hear and see.