电子科技大学：《图像处理及应用 Image Processing and Application》课程教学资源（课件讲稿）Chapter 08 Image Compression

电子科枝女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC Chapter08 mage Compression Ping Zhang

Ping Zhang

电子科发女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC Outline ◆ Background ◆Fundamentals Some Basic Compression Methods Digital Image Watermarking' 河

Outline  Background  Fundamentals  Some Basic Compression Methods  Digital Image Watermarking*  Background  Fundamentals  Some Basic Compression Methods  Digital Image Watermarking*

电子科发女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC Outline ◆ Background ◆Fundamentals Some Basic Compression Methods Digital Image Watermarking' 河

Outline  Background  Fundamentals  Some Basic Compression Methods  Digital Image Watermarking*  Background  Fundamentals  Some Basic Compression Methods  Digital Image Watermarking*

电子科发女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC 8,1 Fundamentals Agenda g Coding Redundancy g Spatial and Temporal Redundancy Irrelevant Information g Measuring Image Information Fidelity Criteria Image Compression Models Image Formats,Containers,and Compression Standards Agenda

 Coding Redundancy  Spatial and Temporal Redundancy  Irrelevant Information  Measuring Image Information  Fidelity Criteria  Image Compression Models  Image Formats, Containers, and Compression Standards Agenda 8.1 Fundamentals

电子科枝女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC 8,1 Fundamentals ● The term data compression refers to the process of reducing the amount of data required to represent a given quantity of information ·Data≠Information Various amount of data can be used to represent the same information Data might contain elements that provide no relevant information data redundancy

8.1 Fundamentals • The term data compression refers to the process of reducing the amount of data required to represent a given quantity of information • Data Information • Various amount of data can be used to represent the same information • Data might contain elements that provide no relevant information : data redundancy 

电子科技女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC 8,1 Fundamentals ■ Let b and b'denote the number of information carrying units in two data sets that represent the same information The relative redundancy R is define as: R=1-1/C where C,commonly called the compression ratio,is C=b/b

 Let b and b’ denote the number of information carrying units in two data sets that represent the same information  The relative redundancy R is define as : where C, commonly called the compression ratio, is R 1 1/  C 8.1 Fundamentals ' C bb  /

电子科发女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC 8,1 Fundamentals ■Ifb=b'C=1andR=0 no redundancy ■Ifb>b'C→and R→1 high redundancy ■Ifb undesirable In Image compression,3 basic redundancy can be identified 。Coding redundancy Spatial and Temporal Redundancy Irrelevant Information

 If b = b’ , C = 1 and R = 0 no redundancy  If b >> b’ , C and R high redundancy  If b << b’ , C and R undesirable  In Image compression , 3 basic redundancy can be identified  Coding Redundancy  Spatial and Temporal Redundancy  Irrelevant Information   1  0   8.1 Fundamentals

电子科发女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC 8,1 Fundamentals ab c FIGURE 8.1 Computer generated 256 X 256 X 8 bit images with(a)coding redundancy,(b)spatial redundancy, and (c)irrelevant information.(Each was designed to demonstrate one principal redundancy but may exhibit others as well.) Coding Redundancy Spatial and Temporal Redundancy Irrelevant Information

 Coding Redundancy  Spatial and Temporal Redundancy  Irrelevant Information 8.1 Fundamentals

电子科技女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC 1.Coding Redundancy Recall from the histogram calculations 卫() k=0,1,2,L-1 MN p,(r)is the probability of a pixel in M XN image to have a certain value rk; nk is the number of times that the kth intensity appears in the image; L is the number of intensity value. If the number of bits used to represent r is I(r),then L = ∑1)p,) k=0

 Recall from the histogram calculations pr(rk) is the probability of a pixel in M X N image to have a certain value rk ; nk is the number of times that the kth intensity appears in the image; L is the number of intensity value.  If the number of bits used to represent rk is l(rk), then ( ) 0,1, 2..., 1 k r k n pr k L MN    1. Coding Redundancy 1 0 () () L avg k r k k L lr p r    

电子科技女学光电科学与工程学院 SCHOOL OF OPTOELECTRONIC SCIENCE AND ENGINEERING OF UESTC Examaple8.1 A simple illustration of variable-length coding Tk p() Code 1 1i(rk) Code 2 12(k) r87=87 0.25 01010111 8 01 2 128=128 0.47 10000000 8 1 1 186=186 0.25 11000100 8 000 3 255=255 0.03 11111111 8 001 3 rk for k≠87,128,186,255 0 8 0 4 256×256×8 1)P) C= ≈4.42 k=0 256×256×1.81 =2(0.25)+1(0.47)+3(0.25)+3(0.03) =1.81bit R=11 =0.774 4.42

4 0 () () 2(0.25) 1(0.47) 3(0.25) 3(0.03) 1.81 bit avg k r k k L lr P r       1 1 0.774 4.42 R   Examaple8.1 A simple illustration of variable-length coding 256 256 8 4.42 256 256 1.81 C      