DIGIT VIDEO SYSTEMS A. ASTAPKOVICH State University of Aerospace Instrumentations, Saint-Petersburg, 2012 Lecture 0 COURSE REVIEW.

DIGIT VIDEO SYSTEMS

A. ASTAPKOVICH

State University of Aerospace Instrumentations , Saint-Petersburg, 2012

Lecture 0

COURSE REVIEW

GOAL OF THE COURSE

Establishing the understanding of the basic principals of digit video systems :

digit video processing cycle;

structure of the modern digit video systems ;

applications in multimedia, industrial, security, research ; review of the modern research activities ;

Course structure

PHYSICAL PRINCIPALS OF THE DIGIT FRAME FORMING AND HARDWARE/SOFTWARE REALIZATIONS

MODERN APPLICATIONS: MULTIMEDIA STANDARTS,

DIGIT VIDEO FOR INDUSTRY AND SPACE RESEARCH

DIGIT VIDEO PROCESSING BASICS

REVIEW OF MODERN RESEARCH ACTIVITY

TOPIC 1. PHYSICAL AND REALIZATION BACKGROUD

Lecture 1. DIGIT CAMERA § 1. Digit frame forming § 2. Noise and distortions in digit video § 3. Low level processing § 4. High level processing

§ 5. Structure of the modern digit video camera

Lecture 2. DIGIT VIDEO § 1. Digit video in multimedia

§ 2. Multimedia standards review § 3. Digit video in industry and space applications § 4. Modern silicon solutions § 5. System architecture and high level software

DIGIT CAMERA STRUCTURE

Lecture 3-4. IP-VIDEOPHONE AND SECURITY SYSTEMS §1. Packet communication nets §2. Family of IP-videophones CISCo and basics multimedia

standards §3. Family of IP-videophones CISCo and basics multimedia standards

§4. Audio codec and the quality of the videophones §5. Multichannel security video systems

TOPIC 2. DIGIT VIDEO BASED SYSTEMS

Cisco IP phone 7985 Videophone Digital Media System-on-Chip(DMSoC) TMS320DM365

Lecture 5-6. TECHNICAL VISION SYSTEMS §1. Digit video systems for industry applications

§2. Review of the mars rover Spirit-Opportunity control system

§3. Mars rover video system §4. Video in the mars rover control system loop

§5. Special features of the digit channel §6. Space standards of ECSS

Mars rover Spirit-Opportunity

APPLICATION EXAMPLES AND SPACE STANDARTS

TOPIC 3. DIGIT IMAGE PROCESSING

Lecture 7. DIGIT VIDEO COMPRESSING BASICS §1. Video stream parameters and and the image quality

estimation §2. Video compression basic ideas §3. Wavelet compression and Haar basis §4. Review of the wavelet compression algorithms

HAAR wavelet1 0 ≤ x ≤ ½

Ψ(Ψ(xx) =) = -1 ½ ≤ -1 ½ ≤ x x ≤ 1≤ 1 0 1 ≤ 0 1 ≤ x; x; x x ≤ ≤ 00

HAAR basisΨ Ψ j ij i((xx) ) = = Ψ (Ψ (22jj x - i x - i) ) i= 0..2 i= 0..2 j j -1-1

MASTER_PIC MH1

(a+b+c+d)/4 (a-b+c-d)/4

(a+b-c-d)/4 (a-b-c+d)/4

LL LH

HL HH

W0

V1

V2

Lecture 8-9 IMAGE PROCESSING §1. Image quality estimation §2. Image filtering algorithms §3. Edge detectors §4. Moving object extractions

IMAGE PROCESSING

Original image Enhanced contrast

PSNR=25 dB

JPEG compression

PSNR=25 dB

EDGE DETECTORS

CLEAN AMAGE

NOISY IMAGE

IMAGE EDGES

EDGE

DETECTOR

CANNY

DIGIT VIDEOSTREAM PROTECTION

Lecture 10. STEGANOGRAPHY § 1. Basic definitions and digit watermark classifications § 2. Digit watermark system structure § 3. System and algorithm requirements § 4. Attack review § 4. Application examples

 Encryption keys -

K Container - I

• image• audio sample• text• code

Marking information - M•trade mark• copy number• other

EMBEDDINGALGORITHM

Marked information

or STEGO - I”

LL LH

HL HH

Q=55

COMPRESSION JPEG,

JPEG2000

WATERMARKEXSTRACTIO

N

APPLICATION EXAMPLE

Modified Kutter algorithm

MODERN APPROACHES

Lecture 11. ADAPTIVE ALGORITM PARADIGM § 1. Modern structural wavelet based norms SSIM and CW-SSIM

§ 2 Neuron net based algorithms § 3 Adaptive boosting learning ( ADA BOOST)

Original image

MSE=0

SSIM=1

CW-SSIM=1

MSE=306

SSIM=0.928

CW-SSIM=0.938

Enhanced contrast Distorted brightness

MSE=309

SSIM=0.987

CW-SSIM=1

Gauss noise

MSE=309

SSIM=0.576

CW-SSIM=0.814

Noise proof edge detector on base of ANN

S1(0,0) S2(0,0)

S1(0,1) S2(0,1) ................. 1……………………………….. S1(i,j) S2(i,j ) Snsen(i,j) 1

w1

w2

wnsen+1

F(0,0)

F(0,1)

………F( i,j)

W = (ST S + E) –1 ST F

min F(w) = (SW - F, SW – F)+ (W,W) w

S * W = F

NNEdgefilter

MASTER_PIC