KR20000022960A - Methods and arrangements for converting a high definition image to a lower definition image using wavelet transforms - Google Patents

Abstract

PURPOSE: A system and a method for converting a high definition image into a low definition image using wavelet transform are provided to supply a low definition digital video signal from a high definition digital video signal and to convert the high definition digital video signal into the low definition digital video signal for display. CONSTITUTION: A signal conversion algorithm is used which enables more effective signal conversion, the signal conversion algorithm is used in transmitting/coding or receiving/decoding in order to generate a more intense and accurate output. This signal conversion algorithm improves the function by being applied to a conventional system, and a complete transmission and reception system can be designed where the signal conversion algorithm is applied to data coding, decoding and decimation operation. A system(100') includes a conventional encoder(102) applying a DCT(Discrete Cosine Transform) algorithm to image data (114) and transmits the encoded signal through a transmission link(106).

Description

Methods and arrangements for converting a high definition image to a lower definition image using wavelet transforms}

TECHNICAL FIELD The present invention relates to multimedia video and digital communications, and more particularly, to an apparatus and method for converting a high definition image or an image into a lower definition image.

Many different video and / or video sampling techniques are used for coding, transmitting and playing back signals and / or multimedia images, such as, for example, still and moving pictures, other related data signals such as video and audio. These techniques allow multimedia information to be properly encoded, transmitted and reproduced by known hardware in use at present. Examples of such techniques are well known in the prior art, and many of them are described in ITU-T Recommendation H.262 ISO / IEC 13818-2: 1995, Information Technology (Comprehensive Encoding of Moving Pictures and Related Audio Information Since March 31, 1995: Video).

On November 4, 1994, the International Organization for Standards (ISO) Motion Picture Experts Group (MPEG) adopted a standard for audio / video digital compression known as MPEG-2 (MPEG-2). This standard recognizes consistent digital signal sampling, coding, transmission and reception worldwide and is well known in the art.

U.S. Patent No. 5,262,854, issued to Ng on November 16, 1993 under the title 'Low Resolution HDTV Receiver', was extracted to provide compressed HDTV digital video signal data in lower resolution NTSC images (decimation). Looks receiver. This system allows high definition signals to be used in lower definition receivers, which are now more commonly used than high definition receivers.

Similarly, many other types of video sampling techniques and digital component video formats are commonly used for MPEG video coding. For example, there is a high definition video format which defines 4: 4: 4 as the relative relationship between luminance and color components in a transmitted digital video color signal. Lower definition video sampling formats such as 4: 2: 2 and 4: 2: 0 contain fewer color components per luminance sample of the digital signal. All three sampling techniques are well known in the art. Higher definition sampling techniques contain more information and thus provide higher resolution images.

Regardless of the sampling technique, suitable display devices such as a monitor or flat panel display are required to effectively reproduce the encoded image. Despite current developments in high resolution systems and devices, displays capable of reproducing and displaying higher resolution images will be very expensive. For example, high definition television (HDTV) devices can cost thousands of dollars. For many consumers, the cost of HDTV will seem unprecedentedly expensive compared to the cost of standard definition televisions such as NTSC compatible devices that cost less than a few hundred dollars.

Similar cost issues exist for video signal broadcasters and producers. Generating higher resolution images requires a level of art image recording and generation system, and often requires additional channels supplied within the transmission channel to control the growth of information (data) to be supplied to consumers. .

Broadcasters and consumers are also thinking that not many consumers will have high-resolution display devices. This is particularly the transition as the next generation of video technology that will incorporate HDTV as a standard.

Thus, a method and apparatus that enables the remaining consumers with lower definition televisions and imaging devices to receive higher definition image data and replace it with lower definition image data that can be displayed on lower resolution displays. Is needed.

HDTV digital video signal decoders are also well known in the art. Conventional MPEG compatible decoders generally have an inverse discrete cosine transform (IDCT) process that is used to decode video related data previously encoded using a discrete cosine transform (DCT) process.

Image data encoded (encoded) / decoded (decoded) by a conventional encoder or decoder generally contains three components per pixel. The components are luminance data Y _c , chrominance data U _c and chrominance data V _c . For example, in order to display a high definition image such as a 1920x1080 pixel image, a general decoder must output 1920x1080 pixels of luminance-related data and 960x540 pixels of color-related data. In this example, the final data provides a 4: 2: 0 image with 1920x1080 pixels.

In order to decimate or otherwise reduce the amount of image data, known methods and devices attempt to create a subset of the image data so that it can be displayed on a low resolution display. In order to achieve this "downscaling", known methods and devices usually preselect or filter the received encoded image data. For example, these methods use masking techniques to remove specific data. The remaining amounts of encoded image data are then decoded using a decoder or the like having an IDCT process. Decoded image data is now filtered and / or decimated to further reduce the image to be displayed on lower resolution displays.

For example, in some decoders the amount of information used for low-definition images is one-quarter of the amount of information used for the original higher definition images. Thus, a low resolution image for a 1920x1080 pixel high definition image is 960x540 pixels.

It should be noted that known decoders of this kind inevitably lose video-related information before and after the IDCT process. One consequence of losing video-related information is that it can adversely affect the symmetry of the final decoded image. The loss of symmetry of the final decoded image from this kind of known decoder can result in lower sharpness images, for example, lower resolution images that are asymmetrical 4: 2: 0.

1 and 2 show a typical digital video encoding / decoding transmission system block diagram. 1 is a block diagram depicting a conventional system 100 having an encoder 102 that encodes an image file 104 that includes image data 114. The output of the encoder 102, i.e., the encoded image data, is transmitted to the decoder 108 via the transmission link 106 or otherwise supplied.

The transmission link 106 includes one or more communication media and / or systems and an auxiliary mechanism configured to carry the encoded image data from the encoder 102 to the decoder 108. Examples of transmission link 106 include, but are not limited to, telephone systems, cable television systems, direct or indirect broadcast television systems, direct or indirect satellite broadcast systems, one or more computer networks and / or buses, the Internet, intranets and software, hardware and other Communication systems and related devices will be included.

The decoder 108 decodes the received encoded image data and outputs an image 110 suitable for reproduction through the display 112. In some conventional systems, encoder 102 and / or decoder 108 each include one or more processors coupled with a memory. The processor (s) respond to computer implemented instructions stored in memories to encode or decode image data 114 as needed. In other conventional systems, encoder 102 and / or decoder 108 include logic configured to encode or decode image data 114 as needed.

2 is a block diagram depicting a conventional encoding / decoding / transmission system 100 that reduces a higher sharpness image 114 to a lower sharpness image 124 that can be displayed on a lower resolution display (not shown). to be. System 100 includes an encoder 102 that builds a DCT algorithm 116 that encodes image data 114 using a DCT algorithm. Decoder 108 then calculates the encoded image signal using pre-parser algorithm 118, IDCT algorithm 120, and post filter algorithm 122 in FIG. ) Pre-parser algorithm 118 reduces the amount of encoded image data from encoder 102 in decimation, filtering, masking, and / or otherwise, and IDCTs the subset of encoded image data received for subsequent processing. Output to algorithm 120.

IDCT algorithm 120 then decodes the subset of encoded image data and outputs the decoded image data to post filter algorithm 122. Post filter algorithm 122 further forms the decoded data and forms the reduced image 124.

Post filter algorithm 122 generally downsamples the decoded data in decimation, filtering, and / or otherwise. Reduced image 124 represents a lower sharpness image suitable for display on a lower resolution display. 5 depicts a typical matrix operation 200 associated with the DCT / IDCT algorithm. Matrix D is a 8x8 matrix (i.e., macroblock) image data and this is multiplied by 8x8 DCT / IDCT coefficient matrix C and C ^T provides an 8x8 data matrix T.

The data matrix T of FIG. 5 is eventually provided to the decoder 108 via the link 106. Tables 1a and 1b show a typical computer program that includes an IDCT procedure with an inverse fast discrete cosine transform.

As shown in Table 1A, section 300 is included to indicate the mathematical steps to build an inverse fast discrete cosine transform. The algorithms included in the computer programs in Tables 1A-1B and in particular the coefficients applied to the matrix operation 200 are based on, for example, the DCT and IDCT defined in the referenced sections 304, 308 of Table 3A and 310 of Table 3B.

However, the reduced image 125 was processed to be substantially time consuming and ineffective in order to produce a low quality image.

It is an object of the present invention to provide a system and method for supplying a low definition digital video signal from a high definition digital video signal. Another technical problem to be achieved by the present invention is to provide a system and method for quickly and effectively converting a high definition digital video signal into a low definition digital video signal format for display.

1 is a block diagram depicting a conventional digital video encoding / decoding / transmission system.

2 is a block diagram depicting a conventional digital video encoding / decoding / transmission system that reduces a higher definition image to a lower definition image to be displayed on a lower resolution display.

3 is a block diagram depicting an improved system according to a first preferred embodiment of the present invention for reducing high quality digital images to low quality images to be displayed on a low resolution display.

4A is a block diagram depiction of an improved system according to a second preferred embodiment of the present invention for reducing high quality images to one or more low resolution images to be displayed on different low resolution displays.

FIG. 4B is a block diagram of a system combining IDCT images and wavelet reduced images according to a second preferred embodiment of the present invention shown in FIG. 4A.

FIG. 5 depicts a typical matrix operation associated with the DCT / IDCT process shown in FIG.

6A depicts an explicit matrix operation associated with a typical IDWT / DWT algorithm used by preferred embodiments of the present invention.

6B depicts an explicit matrix operation associated with a typical fast IDWT / DWT algorithm used by preferred embodiments of the present invention.

7 is a block diagram of an enhanced system according to a third preferred embodiment of the present invention in which DWT is used to encode image data for transmission.

8 is an example of an HDTV image file modified using DCT encoding and / or IDCT decoding to supply an image, in accordance with the operation of FIG. 4B.

The present invention relates to a system for converting a high definition image or data to a lower definition image, and in the preferred embodiments used in the present invention, using a predetermined signal switching algorithm that enables more efficient signal switching and is more robust. The above problem is solved by using a signal switching algorithm in both transmission / coding or reception / decoding to generate an accurate output.

This or another object of the present invention is solved by incorporating a wavelet transform to generate coefficients that are part of a discrete wavelet transform (DWT) and / or an inverse discrete wavelet transform (IDWT) within the method and apparatus of the present invention. Can be.

For example, according to the first preferred embodiment of the present invention, the IDWT process decodes and decimates the encoded higher sharpness image data to generate lower sharpness image data suitable for display on a lower resolution display device. It is advantageous to be included in a decoder for this purpose.

According to a preferred embodiment of the present invention, the decoding and decimation of DCT encoded image data is enhanced in the decoding process and easily performed by a decoder having an IDWT process for performing both decoding and decimation. Image data decoded by an IDWT-decoded decoder can be displayed on a lower resolution display as a 4: 2: 0 video image.

This IDWT decoded 4: 2: 0 video image is symmetric because the received encoded image data is not pre-parsed or otherwise filtered prior to being decoded by the IDWT process. Instead, all of the received encoded image data is processed using IDWT. When applied to the received encoded image data, the IDWT process essentially decimates or downsamples the amount of video data. The IDWT uses the coefficients of the IDWT to reduce the capability of one or more wavelet transforms to be applied to discrete blocks of received encoded video data.

An additional benefit of IDWT-decoded decoders is that for pictures such as MPEG-2 images, motion compensation is performed on the decimated output of the IDWT process.

Known decoders usually perform motion compensation on a 16x16 block or matrix of image data. The decoder in which the IDWT is formed reduces, according to the first preferred embodiment of the present invention, a matrix or block of image data to one quarter of its original size, which is 8x8. This block of 8x8 image data is then interpolated to its original size for a while, and motion vectors, such as those commonly used in 16x16 blocks, with reduced number of operations and increased speed are applied. The reduced size of the image data also reduces the memory need of the decoder, such as cache memory supporting one or more processors included in the decoder.

Accordingly, the present invention provides a method and apparatus that can allow a user to receive high definition image data and replace it with data for display on an existing television set or a less expensive high resolution display. The method and apparatus of the present invention may be used by producers and / or signal broadcasters to generate very good high definition image data.

The foregoing and further features, configurations, advantages and objects of the invention, which are considered novel, are described in detail in the appended claims. The construction and operation of the present invention will be fully understood from the following detailed description and accompanying drawings, together with additional objects and advantages. Each of the figures included herein is not to be considered an exact description of an embodiment of the present invention, but is provided for illustrative purposes only and can be interpreted in connection with the accompanying specification.

The following description is provided to enable any person skilled in the art to make or use the invention, and describes the best embodiment presently anticipated by the inventors carrying out the invention. However, it is very apparent that there is room for a variety of modifications to those skilled in the art after the generic principles of the invention have been defined herein.

3 is a block diagram depicting an improved system 100 'in accordance with a first preferred embodiment of the present invention for reducing high definition digital images to low definition images that can be displayed on lower resolution displays. As shown in FIG. 3, the system 100 ′ of the first preferred embodiment of the present invention includes a conventional encoder 102 that applies a DCT algorithm to image data 114, and transmit link 106 as described above. The coded signal is transmitted through

A decoder 108 'made in accordance with the first preferred embodiment of the present invention is connected to the link 106 and is configured to receive encoded image data from the encoder 102. The decoder 108 'applies an IDWT algorithm 132 that is formed to decode and decimate the encoded image data and outputs the wavelet reduced image 134. The output from the IDWT process is a wavelet reduced image 134 indicating an improvement to the reduced image 124 generated by the conventional decoder 108 depicted in FIG. The wavelet reduced image 134 is symmetrical and motion compensated. Moreover, the IDWT algorithm 132 tends to increase the processing speed in the decoder 108 'of the first preferred embodiment of the present invention.

4A is a block diagram of an enhanced system 100 "according to a second preferred embodiment of the present invention. The system 100" of the second preferred embodiment of the present invention may display high definition images on different low resolution displays. Reduce to one or more low-definition images.

The system 100 "of the second preferred embodiment of the present invention is a conventional encoder 102 that applies a DCT algorithm that encodes image data 114 and transmits the encoded data over the transmission link 106 as described above. ).

A decoder 108 " in accordance with a second preferred embodiment of the present invention is connected to the transmission link 106 and is configured to receive encoded data from the encoder 102. The decoder 108 " It is a hybrid (hereinafter hybrid) of the phosphor decoder 108 and the decoder 108 'in the first preferred embodiment of the invention depicted in FIG. As shown, decoder 108 " applies DCT algorithm 120 and IDWT 132, each of which is configured to decode the encoded image data separately.

The IDCT algorithm 120 outputs an IDCT image 136 that represents a high definition image 114 as encoded by the DCT algorithm 116 in the encoder 102. The IDWT algorithm 136 not only decodes the received encoded data, but also decimates the encoded image data (as described above) to generate a wavelet reduced image 134.

The wavelet reduced image 134 and IDCT image 136 in the second preferred embodiment of the present invention are depicted in the block diagram of FIG. 4B. Wavelet reduced image 134 includes Y _W data 134Y, U _W data 134U, and V _W data 134V. According to one embodiment of the invention, the wavelet reduced image 134 provides a 4: 2: 0 video image. IDCT image 136 includes Y _C data 136Y, U _C data 136U, and V _C data 136V. IDCT image 136 according to the second preferred embodiment of the present invention also provides a 4: 2: 0 video image.

As shown, hybrid image 140 may be provided by combining Y _W data 134Y with U _C data 136U and V _C data 136V, in accordance with a second preferred embodiment of the present invention. In the second preferred embodiment (in either embodiment), the hybrid image 140 will result in a 4: 4: 4 video image.

For example, when trying reference to Figure 8, HDTV file 400 having a 1920x1080 pixel MPEG-2 standard Y _C data for each frame in accordance with (that is, DCT encoding and / or using the IDCT decoding process) is 1920x1080 pixels , V _C data is converted 960x540 pixels U and _C data to generate IDCT image 402 of 960x540 pixels. The same HDTV image file 400 is 960x540 pixels in Y _W data and 480x270 pixels in U _W data for each frame according to a second preferred embodiment of the present invention (i.e., using the DCT encoding and / or IDWT decoding process). And V _W data is converted to produce a wavelet reduced image 404 that is 480x270 pixels. As shown, the hybrid image 140 ′ may be created by combining Y _W data of 960 × 540 pixels with U _C and V _C data of 960 × 540 pixels. The combined result, the 4: 4: 4 video image, will have higher clarity than the 4: 2: 0 video image, which is the wavelet reduced image 404.

6A and 6B depict respective verifying matrix operations 202 and 202 'associated with a typical IDWT process used by preferred embodiments of the present invention. In operation 202, the information making up the data matrix T is received from the encoder 102 via the link 106 and multiplied by the 4x8 IDWT coefficient matrix W and the 8x4 IDWT coefficient matrix W ^T to produce a 4x4 matrix dTI.

In another embodiment of the present invention, i.e., in 202 'operation, data matrix T is received from encoder 102 via link 106 and multiplied by 4x7 fast IDWT coefficient matrix W and 7x4 fast IDWT coefficient matrix W ^T to 4x4. Generate the matrix dTI.

Tables 2a and 2b illustrate typical computer programs including IDWT procedures that can be used by preferred embodiments of the present invention.

As shown in Table 2a, section 302 is included to indicate the mathematical steps to build an IDWT process. The algorithms included in the computer programs of Tables 2a and 2b, and in particular the coefficients applied to the matrix operation 200, are related derivations within the reference block 306 of Table 3a and also the referenced sections 312 and 314 in Table 3b. Based on the DWT function defined in

Fast DWT / IDWT matrix operations further reduce the number of mathematical operations required, for example, by eliminating operations associated with rows of coefficients equal to '0' in matrix W as derived in section 312 of Table 3b. Is generated. Fast DWT / IDWT matrix operations can be further adapted by identifying components having rows, columns and / or something in common. For example, all components in the first or top row of the matrix derived in section 312 have the same value (at this resolution), and / or in the third row from the top down, some components are positive. Some are negative, but have the same absolute value. Those skilled in the art will appreciate these operations and other reductions that reduce computer computation time and / or reduce the number of operations required.

In a third preferred embodiment of the present invention, DWT is used to encode image data for transmission. For example, in the system 300 shown in the block diagram of FIG. 7, the DWT functions, algorithms and derivations / coefficients shown in Tables 2A-2B and 3A-3D may be used to encode the image data 114. It is included in the DWT process 302 within the encoder 102 '.

Following transmission over channel 106, the coded image data may include one or more decoders (decoders) to generate a lower sharpness wavelet reduced image 134 and / or a higher sharpness wavelet interpolated image 304. Such as 108 '. In the third embodiment of the present invention, instead of encoding the DCT process, the DWT process is used for both encoding and decoding. Switching to wavelet-based transformations and adapting matrix operations tends to reduce the number of operations required and the communication and / or memory requirements within the overall system. The result is that some different (high or low) sharpness images can be supplied at different display resolutions.

According to some aspects of the present invention, other wavelet transforms may be used within specific systems and / or for any kind of image of the methods and apparatuses of the present invention.

Those skilled in the art will appreciate that various applications and modifications of the above-described preferred embodiments can be shaped without departing from the spirit and scope of the invention. It is, therefore, to be understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein.

In accordance with the present invention, wavelet transforms can be used to switch images of a given sharpness and to adapt matrix operations to reduce the number of operations and communication and / or memory requirements required to switch the sharpness of an image within the overall system.

Claims (13)

Means for receiving the transmitted high definition signal; Decoding means for decoding the received high-definition signal and applying a predetermined algorithm for decoding and decimating the transmitted high-definition signal to generate a reduced image of lower sharpness; And And output means for outputting the reduced image. The method of claim 1, wherein the predetermined algorithm, High-definition video decoder characterized by an Inverse Discrete Wavelet Transform (IDWT) algorithm that generates a wavelet reduced image. The method of claim 2, And a second decoding means for decoding the received high definition signal and generating an IDCT image by applying an IDCT algorithm to the transmitted high definition signal. The method of claim 3, And a means for combining the IDCT image and the wavelet reduced image. Means for transmitting a high definition signal representative of the image data; Coding means for generating a high definition signal in the transmission means by applying a predetermined coding algorithm for encoding the image data; Means for receiving the transmitted high definition signal; Decoding means for decoding the received high definition signal and applying a predetermined decoding algorithm for decoding and decimating the transmitted high definition signal to generate a reduced image of lower sharpness; And And output means for outputting the reduced image. The method of claim 5, wherein the predetermined coding algorithm, A high definition video transmission system characterized by a DWT algorithm. The method of claim 5, wherein the predetermined decoding algorithm, A high-definition video transmission system characterized by an Inverse Discrete Wavelet Transform (IDWT) algorithm that generates a wavelet reduced image. The method of claim 7, wherein the algorithm, A high definition video transmission system characterized by an algorithm for generating wavelet interpolated images. The method of claim 5, wherein the predetermined coding algorithm, A high definition video transmission system characterized by a DCT algorithm. The method of claim 9, And a second decoding means for decoding the received high definition signal and generating an IDCT reduced image. Means for transmitting a high definition signal representative of the image data; Coding means for generating a high definition signal in said transmission means by applying a coding algorithm to image data; Means for receiving the transmitted high definition signal; Decoding means for decoding the received high definition signal and generating an output image by applying a decoding algorithm to the transmitted high definition signal; And An output means for outputting an output image, And the algorithm applied to the signal comprises wavelet transforms. Applying a coding algorithm to the image data to generate a high definition signal; Transmitting a high definition signal representative of the image data; Receiving the transmitted high definition signal; Applying a decoding algorithm to the received high definition signal; And Generating an output image And the algorithm applied to the signal comprises wavelet transforms. Receiving the transmitted high definition signal; Decoding the received high definition signal by applying an algorithm to decode and decimate the transmitted high definition signal to produce a reduced image of lower clarity; Generating a reduced image; And Outputting the reduced image.