JPEG 2000 revealed

Abstract As soon as "JPEG" is mentioned, people think of "file format." In fact, it is an abbreviation for "Joint Photographic Experts Group" (Standard Organization for Static Image Coding and Compression).

While people are keen to develop compression technologies, JPEG has developed a file format with its advantages in the standard compression technology. This latest compression technique is documented in ISO 15444 and is called JPEG2000.

Background In the past decade, digital image storage and transmission technologies have been rapidly developed and their applications have become increasingly widespread. Digital cameras and mid-range and low-end scanners are making rapid use of digital images in consumer and commercial applications. At the same time, the massive use of digital images on the Internet and low-cost multimedia computers has also made digital images a part of everyday life for many people.

In addition to these factors, the popularity of color printers also requires the quality of the pictures to meet the high quality printing requirements. The size of an image file meeting these requirements becomes a challenge for image storage and transmission. Therefore, the compression of digital images is critical to the consumer, business and scientific fields.

Current JPEG compression (different from JPEG 2000) is based on the 1988 JPEG committee's use of DCT technology in combination with ISO 10918-1 Hufmann encoding and is now the first of several static image compression standard versions.

DCT is a lossy compression method in which the user specifies the image quality and depends on the compression rate of the actual encoded image. It quantizes the depth of the pixel as a parameter to complete the conversion of the bitmap. A low-quality image using Hufmann coding can achieve a good compression ratio at a certain quantization level. These technologies allow multiple images to be encoded on screens and printers with good quality.

In 1988, the use of DCT-computing techniques for image applications on PCs and workstations was very complicated. Collaborating on different standards, JPEG has a difficult start.

In 1992, C-Cube's CEO served on the JPEG committee to develop a standard file format JFIF with limited colors, algorithms, and other aspects. This technology was immediately accepted by the public as the only JPEG solution on the market. The popularity of JFIF is like the current JPEG file. However, some of its limitations make it difficult to develop in high resolution, true color, and some additional features.

From the very beginning of the JPEG and JFIF applications, some other file formats and compression technologies based on DCT technology began to form, including SPIFF, which is a multi-channel called JTIP format. These have the same implementation called Flashpix.

JPEG 2000
With the increase in computing power of standard computers in 1995, the decoding speed of software exceeded hardware. In 1996, the JPEG organization decided to develop a more compact image format. Especially in the development of fragmentation and waveform compression technology, a new standard named JPEG2000 emerged in 2000.

The ultimate goal is to establish a more advanced image coding system for the new century. This coding system will have the following characteristics:

1. Abandon the current address domain to find a more effective way;
2. Development now does not use a compressed image market;
3. Specifies a development system for image applications.

Different applications have different requirements.

E.g:

â–  Internet applications require fast decoding capabilities in quality and resolution;
â–  Mobile applications require high fault tolerance, low power consumption and advanced decoding, and require high image security and watermarking in e-commerce.
â–  Digital camera compression reduces complexity and improves efficiency;
â–  Color fax, printing and scanning require webbing processing;
â–  Digital library and archive applications require the management of data and topics;
â–  Remote sensing, multiple documents, fast encoding, and localized encoding are all important;
â–  Medical applications have a wide range of applications for lossless and lossy coding.

The promotion of this technology has led to more than 20 proposals, most of which are for waveform technology. Adding DCT encoding in this technology is still in the testing phase, but JPEG2000 requires too much and adding DCT is not suitable.

Waveform technology was finally added to JPEG2000, which has the ability to be upgraded both in resolution and quality. Although the complexity of waveform coding depends on the size of the filter and the application of floating-point filters, it is estimated that the waveform coding is more complicated than the current block-based DCT coding. For an entire frame change, the memory requirements of the waveform are greater than the DCT.

Regardless, the line-based implementation can reduce memory requirements. Waveform transforms are currently more complex than DCTs, but as the capabilities of computers continue to increase, waveform transforms are no longer more complicated than DCTs when new JPEGs are released. The JPEG committee believes that hardware requirements for code decoding later are no longer an obstacle to waveform transformation.

Protocol scalability means that image encoding can have different levels of encoding within a stream. Therefore, an identical image file can be applied to different programs. When the image is compressed, it can be online mode, size accumulation or protocol decompression.

Quality may be a quantified problem with resolution and color depth. Users can specify the level of compression, resolution, or quality required when encoding. The file's scalability in quality can be decoded into the required quality just like lossless compression.

These techniques have some interesting features: lossless encoding of images and streaming applications. These capabilities are very useful in the distribution of medical images and programs.

JPEG 2000 reviews all parts of the ISO 15444 standard. The following seven parts are the contents of the plan:

The first part: JPEG2000 image coding system;

The second part: expansion;

Part 3: Motion JPEG 2000;

Part IV: Consistency;

Part V: Reference Software;

Part 6: Develop the image file format;

Part VII: The first part of the support.

The first part is expected to pass ISO standards by the end of 2000. At the same time, the core compression technology and the smallest file format are defined. This format uses a scalable file format structure. Although it contains only the minimum features, it can meet the 80% market demand.

The second part expands the compression and file formats. The basic file is JP2, which is an additional part of this standard, but the implementation of any one format must meet the standard and can read JP2 files. Documents containing extended content do not work in the read-only first part of the program. Users who want to use the second part of the file can use an extended program to read it.

The JPEG2000 code stream is the same for all applications and does not require attention to the file format. The JP2 file format is the default format for applying JPEG2000 programs.

It is worth noting that the compression and file formats are separate. When the program wants to transfer the compressed data in a standard file format, the compression program will run separately.

In many cases, considering the versatility, the file format used to compress image data becomes crucial! The JP2 file format was generated under the cooperation activities initiated by DIG2000. At this event, the JPEG2000 committee was formed by the alliance including the JPEG committee. Their task is to define an image file format that is widely used by general users. They want hardware and software vendors to develop products that use JPEG2000 compression technology for online image databases and web applications.

A more versatile format is increasingly attracting attention. It penetrates the range from TIFF-based to Apple-based QuickTime. This will make choices in different complexity and effects. Finally, part of Apple's QuickTime and MPEG-4 were selected as the file format. The ICC, XML, JPEG2000 compression and image meta formats were referenced when creating this format.

The JPEG2000 Committee recognizes that many organizations want to add JPEG2000 encoding to other file formats. Thus, the widespread use of the JPEG2000 format is a greater challenge for the JPEG2000 committee.

The JP2 format defines how to store the specified color. Part2(jpx) adds more complex functionality to part1. In Part1, reading JP2 files requires sRGB and grey-sRGB support. All read JP2 files must support the use of legal ICC standard color interpreter. These standards are to comply with Three-Component Matrix based and Monochrome Input Profiles. It has low adaptability but good collaboration. Other color spaces can be listed in the JPX file, and support in the JP2 file is not necessary. In addition, the second part defines a better registration procedure for color space than others.

The JP2 file format defines a file as a block for storing image data and image metadata. This makes it easier to analyze these blocks or add new blocks.

These standard blocks in the JP2 file include file names, file types, and compatibility information, general header information, palettes, special color spaces, composition sequences, and code streams. In addition to the file name and file type that must be written out and the header information block must be in front of the code stream, there are no additional restrictions. User-defined blocks can be anywhere in the file and any unknown blocks will be ignored.

Additional feature data elements to include. Defining data elements together with image data can increase the value of the entire file and the ability of the system to store images. Because the data element is transmitted along with the image, it has a great relationship with the content of the image. For example, data elements can describe the image author's name, time of creation, copyright, usage, and some parameters.

One of the applications of data elements is IPR (Intellectual Property Rights), which is used to label the author's name, developed copyright, and transmitted copyright.

Image registration. The image registration process has been defined in the original JPEG and is implemented in the fourth part of the JPEG. Similar features are also of interest to JPEG2000. The registration process provides some simple definitions of the image, called the registration area, and is unique.

The registration of JPEG is divided into three sections: 16 areas are country codes, 16 areas are RA areas, and 32 bits are personal codes. This feature is provided by JURA for the world registration code.

Watermark. Protecting image content from piracy and illegal use is a difficult task. Watermarking is a viable solution. It exists in a specified layer of the image and is invisible but can be detected. It is currently intended to use a 64-bit storage watermark, which is transmitted along with the registered information. Watermarking technology is still in the development phase and will soon be tested in JPEG2000.

Region of interest (ROI). As a part of the image, the area of ​​interest has gradually become a trend. In JPEG2000, there are two ways to obtain ROI, active and passive.

In active mode, users are free to define areas of interest. The user stores a high-resolution image, and when the image is requested, the selected specific part is displayed.

In the passive mode, JPEG2000 has a distinctly high resolution area when it is encoded. Users can only display a high quality content in an optional area.

Additional Information For more information, visit: , and here are a lot of articles about JPEG and JPEG2000.