procesarea imaginilor - utclujusers.utcluj.ro/~rdanescu/pi_c1.pdf · exemple de aplicatii...
TRANSCRIPT
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Viziunea artificala
Vizunea artificiala (Computer Vision)?
Viziunea artificiala este un domeniu/disciplina care foloseste metode statistice
care infereaza date/informatie din imagini cu ajutorul metodelor
matematice/geometrice, fizicii, si a teoriei invatarii automate (machine learning)
Se bazeaza pe:
- Cunoasterea profunda a modelului camerei si al procesului de formare al
imaginii pentru a obtine inferente simple de la valorile pixelilor individuali pana
la combinarea informatiei de la imagini multiple pentru a obtine un tot unitar
coerent,
- Impunerea anumitor ordonari asupra unor grupe de pixeli pentru ai separa
intre ei sau pentru a infera informatia de forma si a recunoaste obiecte pe baza
trasaturilor geometrice.
Alte denumiri
• analiza de imagini (image analysis)
• analiza scenei (scene analysis)
• interpretarea imaginilor (image understanding)
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Viziunea artificala
Disipline conexe
• Inteligenta artificiala (artificial intelligence)• Robotica (robotics)• Procesarea semnalelor (signal processing)• Recunoasterea de forme (pattern recognition)• Teoria controlului (control theory)• Psihologia (psychology)• Neurostiintele (neuroscience)
Subdomenii:
- Procesarea imaginilor- Recunoasterea formelor- Fotogrametria
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Procesarea imaginilor
Procesarea imaginilor (Image Processing)
- Se ocupa cu studiul proprietatile imaginilor si cu transformarea acestora(imaginilor)
- Majoritaea algoritmilor de viziune artificala necesita procesareaimaginilor
Example de metode:
• imbunatatirea calitatii imaginilor (image enhancement) – printransformarea imaginilor: punerea in evidenta a detaliilorascune/obscure, a trasaturilor de interes
• compresia (reprezentare compacta a imginilor/secventelor pentrutransmisie)
• restaurarea (eliminarea elementelor de degradarecunoscute/modelabile)
• extragerea de trasaturi (localizarea anumitor sabloane – ex: muchii)
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Viziunea artificiala
Domenii de cercetare:
• Detectia de trasaturi (Feature Detection)
• Representarea contrurelor (Contour Representation)
• Analiza imaginilor de profunzime (Range image analysis)
• Modelarea si reprezentarea formelor (Shape modeling and representation)
• Stereo viziunea (Stereo vision)
• Viziunea color (Color vision)
• Analiza miscarii (Motion analysis)
• Visunea active (Active/Purposive vision)
• Invarianti (Invariants)
• Detectia obiectelor (Object detection)
• Recunoastera obiectelor 3D (3D object recognition)
• Aritectura sistemelor de viziune (Vision architectures)
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Viziunea artificiala
Domenii de aplicare
• Inspectie industriala / controlul calitatii (Industrial inspection/quality control)
• Inginerie inversa (reverse engineering)
• Supraveghere si securitate (Surveillance and security)
• Recunoastera fetei (Face recognition)
• Recunoasterea gesturilor (Gesture recognition)
• Monitorizarea traficului (Road monitoring)
• Aplicatii spatiale (Space applications)
• Analiza imaginilor medicale (Medical image analysis)
• Realitate virtuala, teleprezenta si telerobotica) (Virtual reality, telepresence, and telerobotics)
• Vehicule autonome (Autonomous vehicles)
• Cartografiere automata, achizitie automata de modele (Automated map making, model acquisition)
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Viziunea artificiala
Date de intrare
- Imagini captate cu dspozitive de achizitie adaptate pentru intregulspectru frecventa al undelor eletromagnetice
- Imagini din spectrul vizibil – cele mnai folosite (accesibile)
- Alte surse de imagini: unde acustice, ultrasonice (ecografii)
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Viziunea artificiala
Date de intrare
Reflectivitatea laser
Medium IR (Thermal)
Reflectivitatea RADAR
Spectrul vizibil
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Preprocesarea de imagini medicale (imbunatatirea calitatii)
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Segmentare si analiza de imagini medicale
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Recunoastere de tesuturi prin analiza texturii din imagini
medicale
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Procesarea informatiei elastografice
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Detectie obiecte si drum in scenarii de trafic
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Exemple de aplicatii
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Lenses distortion correction
Left - 2D detection error: Undistort vs. Distort
-8.000
-6.000
-4.000
-2.000
0.000
2.000
4.000
6.000
1 2 3 4 5 6 7 8 9
Target no.
Err
or
[pix
els
[
ErrX
ErrY
8.5 mm lens, CCD camera
Distorted imageUndistorted image
Difference image
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Lenses distortion correction
Distorted imageUndistorted image
Right - 2D detection error: Undistort vs. Distort
-1.000
-0.500
0.000
0.500
1.000
1.500
2.000
2.500
1 2 3 4 5 6 7 8 9 10
Point
Err
or
[pix
els
]
ErrX
ErrY
16 mm lens, CCD camera
Technical University of Cluj Napoca
Computer Science DepartmentIMAGE PROCESSING
Bibliografie curs
R.C.Gonzales, R.E.Woods, "Digital Image Processing-Second Edition", Prentice Hall,
2002.
E. Trucco, A. Verri, “Introductory Techniques for 3-D Computer Vision”, Prentice Hall,
1998.
W.K. Pratt, Digital Image Processing: PIKS Inside, Third Edition. 2001 John Wiley & Sons,
Inc.
G. X.Ritter, J.N. Wilson, Handbook of computer vision algorithms in image algebra - 2nd
ed, 2001 CRC Press.
A. Koschan, M. Abidi, Digital Color Image Processing, Wiley & Sons, 2008.
D. Forsyth, J. Ponce, Computer Vision. A Modern Approach, Prentice Hall, 2002.
L. G. Shapiro, G. C. Stockman, Computer Vision, Prentice Hall, 2001
Bibl. UTCN:
S.Nedevschi, "Prelucrarea imaginilor si recunoasterea formelor", Ed. Microinformatica,
1997.
Scot E, Umbaugh, “Computer Vision and Image Processing”, Prentice Hall, 1998.
Internet:
Milan Sonka, V. Hlavac, R. Boyle, “Image Processing, Analysis, and Machine Vision”,
Brooks and Cole Publishing, 1998.
http://www.icaen.uiowa.edu/~dip/LECTURE/lecture.html
Delft : http://www.ph.tn.tudelft.nl/Courses/FIP/frames/fip.html
⇒ ∞ etc.