1、1. Image(图像) :(1 )Picture, Photograph (图片)Visual data (可视数据)Usually two- or three-dimensional (常为二维或三维) ;(2)书本概念:an image may be defined as a two-dimensional function f(x,y),when x and y are spatial(plane) coordinates,and the amplitude of f at any pair of coordinates(x,y)is called the intensity or g
2、ray level of the image at that point(一幅图像可以定义为一个二维函数 f(x ,y)x 和 y 均为空间(平面)坐标,在任何一对空间坐标中的幅值 f 称为图像在该点的强度和灰度)。2. digital image(数字图像):(1)An image which is“discretized,”i.e.,defined on a discrete grid(ex. scanner)(一个被离散化的图像,即在一个离散的网格(例如,扫描仪) 上定义) ;Two-dimensional collection of light intensity values (or
3、 gray values)(二维光强度值( 或灰度值)的集合)Matrix, two-dimensional function(矩阵,二维函数) (2)书本概念:When x, y,and the intensity values of f are all finite,discrete quantities(当 x,y 和灰度值 f均为有限离散数值时,称为数字图像) 。3. digital image processing(数字图像处理):process digital image by means of a digital computer(借助于数字计算机来处理数字图像) 。4. pix
4、el(像素):denote the elements of adigital image(描述数字图像元素) 。5. example of field that use digital image processing(使用数字图像处理领域的实例)见课本 P29。6. fundamental steps in digital image processing(数字图像处理的基本步骤):Image Acquisition(图像获取) image enhancement and restoration(图像增强与恢复)image segmentation(图像分割)feature selectio
5、n(特征选择;此过程又称为 description(描述) )Image Representation(图像表示)image interpretation(图像解释;此过程也可称为 image recognition(图像识别) ) 。7. structure of the human eye(人眼结构) (成像原理)见 P58。8. The colors that humans perceive in an object are determined by the nature of the light reflected from the object(人体感受一个物体的颜色有物体反射光的
6、性质决定) 。9. monochromatic(or achromatic)light (单色光或无色光): light that is void of color(没有颜色的光) ;the only attribute of it is its intensity or amount(它唯一的属性就是强度或大小) 。10. gray level(灰度级):denote monochromatic intensity(表示单色光的强度) 。11. gray scale(灰度级/灰阶): the range of measured values of monochromatic light fr
7、om black to white(从黑到白的单色光的度量值范围) 。12. chromatic(color)light(彩色光):spans the electromagnetic energy spectrum from approximately 0.43 to 0.79 m(彩色光的能量谱跨越 0.43-0.79m 的范围) 。13. Three basic quantities to describe the quality of a chromatic light source(三个基本量来描述彩色光源的质量):Radiance(辐射):the total amount of en
8、ergy that flows from the light source(从光源流出的能量的总量) ;Luminance (measured in lumens (lm)(流明):gives a measure of the amount of energy an observer perceives from a light source(给出观察者从光源感知到的能量的量度) ;Brightness(亮度): Embodies the achromatic notion of intensity(体现了无色的强度概念) 。14. sampling(采样):digitizing the co
9、ordinate values(对坐标值进行数字化) 。15. quantization(量化):digitizing the amplitude values(对幅值数字化) 。16. image (spatial) resolution(图像(空间)分辨率):the largest number of discernible line pairs per unit distance(每单位距离可分辨的最大线对数量) 。17. intensity resolution(灰度分辨率): the smallest discernible change in intensity level(灰度级
10、中可分辨的最小变化) 。18. some basic relationships between pixels(像素间的基本关系)见课本 P90。应掌握 P93 公式 2.5-1、2.5-2、2.5-3。19. spatial domain(空间域):The aggregate of pixels composing an image(构成图像的像素的聚合) 。Spatial processing(空间域操作):intensity transformation and spatial filter(灰度变换和空间滤波)operate directly on pixels(直接在像素上操作) 。
11、特别提示:灰度变换是对像素进行点操作(point processing) ,而空间滤波是进行邻域操作(neighborhood processing) 。20. basic intensity transformation function(基本的灰度转换函数):image negative(图像反转)log transformation(对数变换)power-law(gamma)transformation(幂律(伽马)变换)piecewise-linear transformation function(分段线性变换函数) 。需要知道各个函数的作用与效果(对应图) ,P129。21. hi
12、stogram(直方图):the histogram of a digital image with intensity levels is in the range 0,L-1 is a discrete function h(rk ) =nk,where rk is the kth intensity value and nk is the number of pixels in the image with intens rk(灰度级范围为0,L-1 的数字图像的直方图是离散函数 h(rk)=nk,其中 rk 是第k 个灰度值,nk 是图像中灰度值为 rk 的像素个数) 。怎么画:h(D
13、)= number of pixels which have the gray level equal to D(D=0,12p-1) (h(D)=灰度等于 D 的像素数)Normalized histogram(归一化直方图): H(D)=h(D)/(MxN) (归一化直方图需要做第二步) 。均衡化、匹配(PPT) 。22. smoothing spatial filter(平滑空间滤波器):used to noise reduction(降低噪声) (1)smoothing linear filter(平缓线性滤波):definition(定义):the output(response)o
14、f a smoothing,linear spatial filter is imply the average of the pixels contained in the neighborhood of the filter mask(平滑线性空间滤波器的输出(响应)是包含在滤波器模板邻域内的像素的简单平均值) 。Application(应用):(会模糊细节) 。 (2)order-statistic(nolinear )filter(统计排序(非线性)滤波器):definition(定义):whose response is based on ordering the pixels co
15、ntained in image area encompassed by the filter,and then replacing the value of the center pixel with the value determined by the ranking result(这种滤波器的响应以滤波器周围的图像区域中所含的像素的排序为基础,然后使用统计排序结果决定的值代替中心像素的值) 。Application(应用):median filter(中值滤波) 用于处理椒盐噪声(salt-and-pepper noise) (以黑白点形式叠加在图像上)十分有效。23. sharpin
16、g spatial filter(锐化空间滤波器):(1)first derivative(一阶微分)特征:must be zero in areas of constant intensity(在恒定灰度区域的微分值为 0) ;must be nonzero at the onset of an intensity step or ramp(在灰度台阶或斜坡处微分值不为 0) ;must be nonzero along ramps(沿着斜坡的微分值不为 0) 。 (2)second derive(二阶微分)特征:must be zero in constant area(在恒定区域微分值为
17、 0) ;must be nonzero at the onset and end of an intensity step or ramp(在灰度台阶或斜坡的起点、终点处微分值不为 0) ;must be zero along ramps of constant slope(沿着斜坡的微分值必须为 0) 。 (锐化图像三种方法都要会,P182 )24. combining spatial enhancement methods(混合空间增强法)见 P19125. frequency domain(频域):26. Fourier series(傅里叶级数): any periodical fu
18、nction can be expressed as the sum of sines and cosines multiplied by aappropriate coefficent(任何周期函数可以表示为乘以适当系数的正弦或余弦的和) 。27. Fourier transform(傅里叶变换):28. DFT:discrete fourier transform(离散傅里叶变换) 。 (还有计算,P257)29. FFT:fast fourier transform(快速傅里叶变换) 。30. characteristics of the frequency domain(频率域的特性)
19、:Impossible to make direct associations between specific components of an image and transform(不可能直接关联图像的特定组件和转换)The slowest varying frequency component (u=v=0) corresponds to the average gray-level of an image(最慢的变化频率分量(u = v = 0)对应于图像的平均灰度级)Move away from the origin of the transform result,the low
20、frequencies correspond to the slowly varying components(远离变换结果的原点,低频对应于缓慢变化的分量)Move away further,the higher frequencies correspond to faster and faster gray-level changes:Edges,noise(进一步移动,较高的频率对应于更快更快的灰度级变化:边缘,噪声) 。31. Lowpass filter(低通滤波器): a filter H(u,v)that attenuates high frequencies while pas
21、sing low frequencies(滤波器 H(u,v) ,其在通过低频时衰减高频) 。32. Highpass filter(高通滤波器): a filter H(u,v)that attenuates low frequencies while passing high frequencies(滤波器 H(u,v) ,其在通过高频时衰减低频) 。33. steps for filtering in the frequency domain(频域滤波步骤):multiply fp(x,y)by(-1) x+y to center its transform(用(-1 ) x+y 乘以
22、fp(x,y)移到其变换中心) ;compute the DFT,F(u,v),of the image from step1(计算来自第一步的图像 DFT)generate a real,symmetric filter function,H(u,v) ,of size PQ with center at coordinates (P/2,Q/2),form the product G(u,v)=H(u,v)F(u,v)(生成一个实的、对称的滤波函数 H(u,v),其大小 PxQ,中心在(P/2,Q/2). 用矩阵形成乘积 G(u,v)=H(u,v)F(u,v)) 。obtain image
23、 gp(x,y),where the real part is selected in order to ignore parasitic complex components that we are dealing with padded arrays(得到图像 gp(x,y),其中,为忽略由于计算不准确导致的寄生复分量,选择了实部。Tips:下标 p 指出我们处理的是填充后的阵列)obtain the final processed result,g(x,y),by extracting the MxN region from the top,left quadrant of gp(x,y
24、)(通过 gp(x,y)的左上象限提取 MxN 区域,得到最终处理结果 g(x,y)) 。34. 三种低通滤波器的特点:ideal lowpass filter(理想低通滤波器)butterworth lowpass filter(巴特沃斯低通滤波器)gaussian lowpasss filter(高斯低通滤波器)35. 三种高通滤波器的特点:ideal highpass filter(理想高通滤波器)butterworth highpass filter(巴特沃斯高通滤波器) gaussian highpasss filter(高斯高通滤波器)36. 同态滤波(homomorphic fi
25、ltering)步骤:(P311)37. Image restoration(图像恢复):Improve an image in some predefined sense(在某些预定义义下改善图像) 。38. Model of the image degradation/restoration process(图像退化/恢复过程模型)39. noise model(噪声模型) (如何获取?):40. 怎么处理周期噪声:41. restoration in the presence of noise onlyspatial filter(只存在噪声的复原空间滤波) (分析、简答)42. the
26、 ways to estimate the degradation function(估计退化函数的方法):estimation by image observation(图像观察估计)estimation by experimentation(试验估计) estimation by modeling(建模估计) 。43. inverse filter(逆滤波):P373(5.7-1、5.7-2)44. color image processing(彩色图像处理):full-color and pseudocolor processing(全彩色和伪彩色处理) 。45. full-color(
27、全彩色):46. pseudocolor processing(伪彩色处理,false color):consists of assigning colors to gray values based on a specified criterion(基于一种指定的规则对灰度值赋以颜色的处理) 。 (包括)intensity slicing(灰度分层) intensity to color transformations(灰度到彩色的变换)47. color model(彩色模型):the RGB model:red、green、blue model(红、绿、蓝色彩模式)the CMY and CMYK color models:cyan、magenta、yellow model and cyan、magenta、yellow and black model(青色、深红色和黄色色彩模式以及青色、深红色、黄色和黑色色彩模式)the HSI color model:hue,saturation,intensity model(色调、饱和度和强度色彩模式 ) 。