Multimedia Compression

- LZW is a loss less data compression algorithm accepted world wide.

- This algorithm has high throughput potential in implementation.

- The high level algorithm goes like this:

1. Assign all the strings to dictionary.

2. Identify the longest string L available in the dictionary which matches the existing input.

3. Find the dictionary index for L to get the output and remove L from the input.

4. Concatenate L by the next symbol of the input to the dictionary.

5. Repeat the process from step 2.

- OptimFROG is one of the lossless audio compression scheme.

- The goal of OptimFROG is to reduce the size of an audio file to the maximum.

- It permits bit identical restoration for all inputs.

- It is like ZIP type of compression and specialized for audio data compression.

- OptimFROG obtains asymptotically the best lossless compression ratios for audio files.

- It can be plugged-in with Windows, Linux, Mac operating systems for Windows Media Player, Foobar2000, Winamp2/3/5, XMPLay, QCD players.

- It supports for all integer PCM wave formats up to 32 bits.

- A technique that converts analog to digital and digital to analog signal conversion.

- It is to transmit voice information.

- It encodes the differences between successive samples into n-bit data streams.

- The transmitted data is reduced to 1-bit data stream.

Following are the features of Delta Modulation:

- A series segments are used to approximate the analog signal.

- The approximated signal is compared with analog wave for determining the increase or decrease in relative amplitude.

- Only the increase or decrease of the signal amplitude is only sent.

- MPEG-1 is a lossy compression standard for both video and audio.

- It compresses VHS quality raw digital video down to 1.5Mega Bits per Second.

- The compression ration is 26:1

- MPEG-1 introduced MP3 audio format.

- The standard consists of –

1. Systems including synchronization of video, audio and other data.

2. Compressed Video content.

3. Compressed Audio content.

4. Conformance testing.

5. Reference software.

- Vocoder is one of the audio processors

- The characteristic elements of an audio signal are captured by Vocoder.

- These characteristic signals are used to affect the other audio signals.

- It is widely used in speech synthesis.

- A component known as Formant describes the frequency of a sound and noise components that are associated.

- The audio output of Vocoder has synthesized sound that is modulated by a filter.

- The synthesized sound is pulsed to the tempo of the user input voice with the tonal characteristics of the voice.

Following are the desirable properties of speech coders:

- Low Bit-Rate.

- Low Memory Size

- Low Coding Delay

- High Speech Quality

- Robustness Across Different Languages / Speakers

- Robustness in the Presence of Channel Errors

- Good Performance on Non-Speech Signals, like telephone signaling.

- MPEG-2 is one of the international standards for compressing video signals.

- Videos can be compressed prior to transmission or storage.

- The approximate compression ratio is 50:1 of the digital size.

- MPEG-2 standard is widely implemented in digital TV transmission with 200 million decoders.

- MPEG-2 standard exploits the redundancies in the original signal.

Following are the JPEG 2000 features.:

- Compression performance is at low bit-rate.

- Both lossless and lossy compression is performed with a single code stream.

- The transmission in noise environment can be done with low bit-error rate.

- It is applicable for both gray and color images.

- Bi-level imagery, natural imagery and computer generated imagery is applicable for gray and color images.

- It is interactive with MPEG-4 standard.

- Wavelets will be replaced with Discrete Cosine Transform

Following are the major steps that are involved in JPEG compression:

- Discrete Cosine Transformation.

- Quantization.

- Zigzag Scan.

- DPCM on DC component.

- RLE on AC Components

- Entropy Coding.

- Down-sample the factors of resolution by 2 for each dimension. For example - ,Reduce the 640 x 480 resolution to 320 x 240 resolution.

- Code the smaller image with another JPEG mode.

- Decode the image and up the sample encoded image.

- Encode the difference between the decoded image and the original image.

- The above process can be repeated for multiple times.

- This process provides high quality images with low resolution display.

1. Performs the conversion of RGB-color space into YUV color components.

2. The chrominance components U and V can be reduced to half of the pixels in the horizontal direction.

1. Two successive video frames often have small differences.

2. MPEG standard has a standard to reduce this temporal redundancy.

1. It is similar to Fast Fourier Transform.

2. The frame blocks of size 8x8 or 16x16 pixels are represented as frequency components.

3. This transforms into frequency domain by implementing standard formula.

Following are the JPEG operation modes:

- Lossless Mode: Every pixel in the image is encoded even though the compression ratio is low.

- Sequential Mode: A single scan from left-to-right and top-to-bottom compresses the image.

- Progressive Mode: Multiple scans are used to compress the image. The transmission time is long.

- Hierarchical Mode: Multiple resolutions are used to compress the image. Hence the lower resolution is accessed first which does not require the decompressing the whole resolution of the image.