Digital image processing (basics of point a)

图像
Various similar vividness of description or photo of the objective object.

图像分类
Press visibility (visible image, invisible image), by the number of bands (single-band, multi-band, ultra-band), according to space coordinates and luminance continuity (analog and digital).

图像处理
The image series of operations, in order to reach the intended purpose technology.

图像处理三个层次
Narrow image processing, image analysis and image understanding.

图像处理五个模块
Acquisition, display, storage, communications, processing and analysis.

图像数字化
The picture is converted into a form of a computer can be processed - digital image process. It includes two sampling and quantization processes. The position and gray scale of the pixel is the pixel attribute.

采样
The successive images into discrete spatial transform operation is called sampling points. Sampling interval and the sampling aperture size is two very important parameters. Sampling mode: seam, seamless and overlap.

量化
Converting the pixel gray values ​​into discrete integral process is called quantization.

灰度级
Represents a gray level (or grayscale or grayscale value) is referred to as an integer pixel brightness level of a pixel.

图像分类
The digital image can be divided into a difference of gray levels: black and white images, gray scale and color images.

灰度直方图
It reflects the frequency of an image pixel in each gray level appears. Gray scale as the abscissa, the ordinate is the gray level of frequency, plotted with gray level relationship is a frequency histogram of FIG.

采样间隔对图像质量的影响
Generally, the larger the sampling interval, the fewer number of pixels resulting image, the spatial resolution is low, poor quality, and checkerboard international pixel block-shaped in severe; smaller sampling interval, the resulting image more pixels, spatial resolution rate, image quality is good, but the large amount of data.

量化等级对图像质量的影
The more sound quantization level, the richer the resulting image-level, high grayscale resolution, image quality is good, but the amount of data; fewer quantization levels, rich image level due to low grayscale resolution, there will be false contour phenomenon, image poor quality, but the small amount of data. However, in rare cases on the fixed image size reduction can improve the quality of gradation, generating the most likely cause of this situation is to reduce the gray level contrast generally increase the image. For example, the details of rich digital images.

直方图的性质
1) reflect only the gray histogram distribution of an image, but does not reflect the position of the image pixels, that is missing
the location information of the pixel.
2) a histogram of the image corresponds to a unique, but not vice versa. Different images may correspond to the same histogram
3) an image into a plurality of regions, and a plurality of regions of the histogram is the histogram of the original image.

直方图的应用
1) for determining whether an appropriate image quantizing
image binarization threshold value 2) for determining
3) calculate the area of the image of the object
4) the image information calculated: entropy H

图像处理基本功能的形式
Single image → single image, multiple images → single image, a single (or multiple) images → numbers or symbols.

对比度
A contrast grayscale image size, contrast ratio = maximum brightness / minimum brightness.

Released three original articles · won praise 0 · Views 66

Guess you like

Origin blog.csdn.net/weixin_43981560/article/details/104593895