The history and basic approaches of Digital image processing

UoG of UESTC level 2017 Zhang Yuchen
The history of digital image processing can be traced back to nearly a hundred years ago. Around 1920, images were first transmitted from London to New York via submarine cables. The first application of image processing is to improve the image quality of submarine cable transmission between London and New York. At that time, image coding was applied. The coded image is transmitted to the destination by submarine cable and then output by special equipment. This is a historic step forward, with the transmission time reduced from more than a week to three hours. In 1950, the Massachusetts Institute of Technology (mit) produced the first computer with graphical display, Cyclone I, which used a cathode ray tube similar to an oscilloscope to display simple graphics. In 1958, calcomp company developed drum plotter. Gerber developed the NC machine tool into a plane plotter. During this period, electronic computers were mainly used for scientific computing, and the graphics equipment of these computers was used only as simple output devices. With the progress of computer technology, digital image processing technology has also been greatly developed. In 1962, Ivan Suzelan, then a PhD student at MIT, successfully developed an epoch-making drawing board course. This is the first interactive drawing system in history and the beginning of interactive computer drawing. Since then, computers and graphics have become more closely linked. In view of Ivan Suzelan’s outstanding contribution to the creation of computer graphics, he was awarded the Turing Award in 1988, the highest award in the field of computer science. In 1964, the Jet Propulsion Laboratory of California, USA, processed a large number of lunar photos sent by the Voyager 7 spacecraft computer to correct various image distortion in the camera on the spacecraft, and achieved obvious results. In the future space and space technology, digital image processing technology has played a huge role. By the end of the 1960s, digital image processing had formed a relatively perfect discipline system. This theory developed rapidly in the 1970s and began to be applied to medical imaging and astronomy. In 1972, American physicist Alan McRade Comeko and British Electrical Engineer Godfrey Newgate Hornsfield invented axial tomography and used it in skull diagnosis. The first X-ray computerized axial tomography device in the world was developed by EMI company. The device is usually called ct, and some algorithms can be used to reconstruct “slice” images by using perceptual data on objects. These images constitute the reconstructed image inside the object, that is, the reconstructed image based on the projection of human head computer. In 1979, Cormaco and hornsfield were awarded the Nobel Prize in Physiology or Medicine for their great role in promoting the development of medical diagnostic technology. Subsequently, in 2003, the Nobel Prize in Physiology or Medicine was awarded to two scientists, Paul Lauterberg, an American chemist, and Peter Mansfield, a British physicist, who made outstanding contributions to the research of medical imaging equipment. The two winners have made pioneering achievements in using magnetic resonance imaging to display different structures. According to the Karolinska Medical College in Sweden, the pioneering work of the two scientists in the field of magnetic resonance imaging represents a major breakthrough in medical diagnosis, treatment and research. In fact, the success of nuclear magnetic resonance is also inseparable from the development of digital image processing. Even today, the noise reduction of MRI image is still a hot research topic in the field of digital image processing. In the development of digital image, another important achievement that must be mentioned is the charge coupling element. CCD was originally invented by scientists Willard Boyle and George Smith of Bell Laboratory in 1969. Like film, CCD can convert optical images into digital signals. At present, the widely used digital cameras, digital cameras and scanners are developed on the basis of ccd. In other words, the digital images we are studying are mainly acquired by CCD devices. Boyle and Smith shared the 2009 Nobel Prize in Physics for their great contributions to the development of ccd. Digital image processing is one of the most popular technologies nowadays. Its shadow is everywhere in our life. It can be said that this is a technology that changes human life every moment.
Digital image processing allows the use of much more complex algorithms, and hence, can offer both more sophisticated performance at simple tasks, and the implementation of methods which would be impossible by analogy means.
Generally speaking, the research contents of digital image processing mainly include the following aspects: 1) image acquisition and output 2) image coding and compression 3) image enhancement and restoration 4) image frequency domain transformation 5) image information security 6) image region segmentation 7) image target recognition 8) image geometric transformation
There are some fundamental methods to carry the digital image processing. The most significant six ways are as follows:

1. Image transformation: Because the image array is very large, directly in the space domain processing, involving a large number of calculations. Therefore, various image transformation methods, such as Fourier transform, Walsh transform, discrete cosine transform and other indirect processing techniques, are often used to transform spatial domain processing into transform domain processing. Not only can it reduce computation, but it can also be processed more efficiently (for example, Fourier transform can be used in frequency domain digital filtering). The reason for that. At present, the newly developed wavelet transform has good positioning characteristics in both time domain and frequency domain, and has a wide and effective application in image processing.
2. Image encoding compression: Image encoding and compression technology can reduce the amount of data describing the image (take the throne), thus saving image transmission, processing time and storage capacity. Compression can be achieved without truth or distortion. Coding is one of the most important methods in compression technology. It is the earliest and most mature technology in image processing technology.
3. Image enhancement and recovery: The purpose of image enhancement and recovery is to improve image quality, such as noise removal, improved image sharpness and so on. Image enhancement does not take into account the cause of image degradation, but only highlights the parts of interest in the image. If the high frequency component of the image is enhanced, the contour of the object in the image will be clear and the details will be obvious. If the low frequency component is enhanced, the influence of noise in the image can be reduced. Image recovery needs to have a certain understanding of the causes of image degradation. In general, the degradation model is established according to the degradation process, and then the original image is recovered or rebuilt by a certain filtering method.
4. Image segmentation: Image segmentation is one of the key technologies of digital image processing. Image segmentation is a meaningful feature that extracts edges and regions from images, and is the basis for further identification, analysis and understanding of images. Although there are many methods of edge extraction and region segmentation, there is not yet an effective method that can be widely applied to all kinds of images. Therefore, the research on image segmentation is still very deep, and it is one of the hotspots of image processing.
5. image description: Image description is a necessary prerequisite for image recognition and understanding. As the simplest two-valued image, its geometric features can be used to describe the characteristics of an object. The general image description method adopts two-dimensional shape description. It has two methods: Boundary description and area description. Two-dimensional texture features can be used to describe special texture images. With the further development of image processing technology, the study of three-dimensional object description has begun, and the methods of volume description, Surface description and generalized cylindrical description are put forward.
6. Image Classification (recognition): Image classification (recognition) belongs to the category of pattern recognition. Its main content is the decision classification of image segmentation and the feature extraction after certain pre-processing (enhancement, recovery, compression). Classical pattern recognition methods are commonly used in image classification, including statistical pattern classification and syntactic (structural) pattern classification. In recent years, the emerging fuzzy pattern recognition and artificial neural network pattern classification have attracted more and more attention in image recognition.