The basic principle and common method of digital image processing

#1, Digital Image processing common methods: Image enhancement, restoration, encoding, compression

1) Image transformation: Because the image array is very large, directly in the spatial domain processing, involving a large amount of computational. Therefore, a variety of image transformation methods, such as Fourier transform, Walsh transform, discrete cosine transform and other indirect processing technology, the processing of the spatial domain to transform domain processing, not only can reduce the computational capacity, but also to obtain more efficient processing (such as Fourier transform can be in the frequency domain of digital filtering processing). At present, the new research wavelet transform has the good localization characteristic in the time domain and the frequency domain, it also has the widespread and the effective application in the image processing.

2) Image Coding compression: Image encoding compression technology reduces the amount of data (that is, the number of bits) that describes an image in order to save image transmission, process time, and reduce the memory capacity consumed. Compression can be obtained without distortion or under permissible distortion conditions. Coding is the most important method in the compression technology, it is the earliest and more mature technology in image processing technology.

3) Image enhancement and restoration: Image enhancement and restoration are designed to improve image quality, such as noise removal, image sharpness, and more. Image enhancement does not take into account the causes of image quality reduction, highlighting the parts of interest in the image. such as strengthening the high-frequency components of the image can make the object contour clear, the details are obvious, such as strengthening low-frequency components to reduce the impact of noise in the image. Image restoration requires a certain understanding of the causes of the quality of the image, generally speaking, according to the quality reduction process to establish a "quality reduction model", and then adopt a filter method, restore or reconstruct the original image.

4) Image segmentation: Image segmentation is one of the key technologies in Digital image processing. Image segmentation is to extract the meaningful features in the image, and its meaningful features include edges and regions in the image, which is the basis for further image recognition, analysis and comprehension. Although many methods of edge extraction and region segmentation have been studied, there is no effective method for the various images. Therefore, the research on image segmentation is still in-depth, which is one of the hot topics in image processing at present.

5) image description: Image description is an essential prerequisite for image recognition and comprehension. As the simplest two value image can describe the characteristics of the object by its geometrical characteristics, the general image description method adopts two-dimensional shape description, it has the boundary description and the region description method. For special texture images, two-dimensional texture features can be described. With the further development of image processing research, three-dimensional object description has been researched, and the method of volume description, Surface description and generalized cylinder description is proposed.

6) Image Classification (recognition): Image classification (recognition) belongs to the category of pattern recognition, the main content is the image segmentation and feature extraction after some preprocessing (enhancement, restoration, compression), and then the decision classification. The classical Pattern Recognition method, statistical pattern classification and syntactic (structure) pattern classification are often used in image classification, and the newly developed fuzzy pattern recognition and artificial neural network pattern classification have been paid more and more attention in the image recognition.

#2, the basic properties of the image

Brightness: Also known as grayscale, it is the color of the light and shade changes, commonly used 0 ~ 255 (from black to white) is indicated. The following three images are different brightness contrasts.

The effect of brightness on image color

Contrast: It is the ratio between black and white, which is the gradation from black to white. The higher the ratio, the greater the gradation from black to white, and the richer the color performance.

Effect of contrast on image color performance

Histogram: Represents the number of pixels in the image that have each grayscale level, reflecting the frequency at which each grayscale appears in the image. Image in the computer storage form, as if there are many points to form a matrix, these points in line with the rows and columns, each point is the value of the image gray value, the histogram is the number of each grayscale in this point matrix occurrence. We can take a look at the grayscale histogram of the following two different graphs:

#3, histogram equalization

Transform an image into another image with a balanced histogram, that is, an image with the same pixel points within a certain gray scale. Here are the graph changes before and after the histogram equalization and the histogram changes:

#4, the addition and subtraction operation of the image

The addition and subtraction operation of two images: the addition and subtraction of the image is the addition and subtraction of the gray value on the storage Rectangle Point column of the image corresponding. Image addition can add the contents of an image to another image, you can achieve two exposures, or a pair of images of the same scene averaging, so that the noise can be reduced. Image subtraction can be used for motion detection or removal of unwanted additive patterns in an image.

Example of an image addition: the operation in the figure is: (a) + (b) = (c)

A	B	C

Example of subtraction of an image: the operation in the figure is (a)-(b) = (c)

#5, the noise of the image

Image noise: As for hearing, we sometimes hear loud noises when talking on the phone, so that we can't hear what the other person is saying. Similarly, for the image, we can clearly see an image, but sometimes the image will have some of the patterns we do not need, so that we can not clearly see a picture, which is the noise of the image.

The commonly used image de-noising method

Commonly used Denoising method: The filter is used to filter the image with noise.

A graph with noise	Figure after arithmetic average filter
Median filtered figure	Noise-Free Diagram

#6, the application of Digital image processing technology

With the development of computer technology, image processing technology has penetrated into every aspect of our life, in which the application of entertainment and leisure has been deeply rooted. The application of image processing technology in entertainment mainly includes: Film special effects, computer games, digital cameras, video broadcasting, digital TV, etc.

Film Special effects production: Since the 1960s, with the gradual use of computer technology in the film, a new film world to show in front of people, this is a film revolution. More and more computer-produced images have been used in the production of film productions. The glamour of its visual effects is sometimes much more than the movie story itself. Now, it's hard to see that there are no computer digital elements in a movie.

Computer video game: Computer video game screen, is one of the fastest development of electronic games in recent years. From 1996 to now, the progress of the game screen can simply be described by leaps and bounds, with the development of image processing technology, many years ago unimaginable images in today has become a common thing.

Digital camera: The so-called digital camera, is capable of shooting, and through the internal processing of the captured scene into a digital format to store the image of a special camera. Unlike a normal camera, a digital camera does not use film, but uses a fixed or removable semiconductor memory to hold the captured image. Digital cameras can be connected directly to computers, televisions or printers. Under certain conditions, the digital camera can also be directly connected to mobile telephones or handheld PC. Since the image is handled internally, the user can immediately check whether the image is correct, and it can be printed or sent out by e-mail.

Video playback with Digital TV: Home theater in VCD, DVD player and digital TV, a lot of video encoding and decoding and other image processing technology, and video codec and other image processing technology development, but also promote video playback and digital TV like high-definition, high-quality development.

The basic principle and common method of digital image processing