12

 From Image Processing to Computer/Machine Vision Low Level Process

 Low, Mid and High-Level processes
Input: Image
Output: Image
Examples: Noise removal, image sharpening
• Low Level
―Image Acquisition, Image Representation
―Image Digitization, Quantization, Filtering
―Morphological Operations, Color Corrections
―Enhancements, De-noising, Sharpening

13
14
 From Image Processing to Computer/Machine Vision Low Level Process
 Low, Mid and High-Level processes
Input: Image
Output: Image
Examples: Noise removal, image sharpening
• Low Level
―Image Acquisition, Image Representation
―Image Digitization, Quantization, Filtering
Mid Level Process
―Morphological Operations, Color Corrections
―Enhancements, De-noising, Sharpening
Input: Image
Output: Attributes
• Mid Level
―Segmentation, Registration, Stitching Examples: Object recognition, segmentation
―Compression, Data Hiding
―Image Restoration, Blur Correction
―Image Transforms

15
16
 From Image Processing to Computer/Machine Vision Low Level Process
 Low, Mid and High-Level processes
Input: Image
Output: Image
Examples: Noise removal, image sharpening
• Low Level
―Image Acquisition, Image Representation
―Image Digitization, Quantization, Filtering
Mid Level Process
―Morphological Operations, Color Corrections
―Enhancements, De-noising, Sharpening
Input: Image
Output: Attributes
• Mid Level
―Segmentation, Registration, Stitching Examples: Object recognition, segmentation
―Compression, Data Hiding
―Image Restoration, Blur Correction
―Image Transforms High Level Process

Input: Attributes
• High Level
―“Make Sense” Output: Understanding
―Feature Detection, Feature Analysis
Examples: Scene understanding, autonomous
―Classification
navigation
―Multi-View Imaging

17
 Robot Navigation

18
 From Image Processing to Machine Vision Low Level Process
 Low, Mid and High-Level processes
Input: Image
Output: Image
• Low Level
―Image Acquisition, Image Representation Examples: Noise removal, image sharpening
―Image Digitization, Quantization, Filtering
―Morphological Operations, Color Corrections
―Enhancements, De-noising, Sharpening In this Mid Level Process
Course
Input: Image
• Mid Level Output: Attributes
―Segmentation, Registration, Stitching
―Compression, Data Hiding Examples: Object recognition, segmentation
―Image Restoration, Blur Correction
―Image Transforms
High Level Process
• High Level
―“Make Sense” Input: Attributes
―Feature Detection, Feature Analysis Some Output: Understanding
―Classification of this
―Multi-View Imaging Examples: Scene understanding, autonomous
as navigation
well

19
 Topics to be covered in this course
o Image processing basics
o Intensity transformations and Histograms
o Image processing in Spatial and Frequency Domain
o Image filtering
o Image restoration
o Morphological operations
o Image segmentation
o Color image / Texture processing
o Image compression
o Image registration
o Feature Extraction (point, edges, lines, salient features)
o Machine Learning Applied for Scene Understanding
— Basic types of classifiers, Performance parameters for evaluation

20
21

Electromagnetic Spectrum and Digital

Images Acquisition
22
23
24
25
26
27
28
29
30
31
Launched in 1990 the Hubble telescope can take
images of very distant objects.

• An incorrect mirror made many of Hubble’s images

useless
• Image processing techniques came to save the day.

32
 Applications of GIS Data Processing and Visualization
 Digital Elevation Models

33
 Manipulation of Satellite Imagery
o Terrain Classification
o Weather Imaging

34
35
36
37
38
 Lip Reading

39
nViso brings power of emotion detection to
banking world

Tired? Angry? Your car knows how you feel

Research at EPFL Lausanne

40
 Physiological Biometrics
Face, IRIS, DNA, Finger Prints

 Behavioral Biometrics
Typing Rhythm, Handwriting, Gait

41
42
43
 Advanced Driver Assistance Systems, or ADAS, to help the driver in
the driving process.
o Designed with a safe Human-Machine Interface .
o To increase car and road safety.

Example systems are,

• Pedestrian collision warning

• Forward collision warning
• Lane departure warning
• Headway monitoring and warning
• Night vision Automatic parking
• Adaptive cruise control

44
45
46
47
 In Short

48
 Conferences
o IEEE ICIP Conference on Image Processing
o IEEE ICCV
o IEEE CVPR Comp Vision and Pattern Recognition
o ACM Sigraph
o (MICCAI) Medical Image Computing and Computer Assisted Intervention

 Journals
o IEEE Transactions on Image Processing
o IEEE Transactions on Medical Imaging
o Elsevier’s Journals
o Springer Journals

 Image Processing Research Groups

o http://peipa.essex.ac.uk/info/groups.html#USA
o http://homepages.inf.ed.ac.uk/rbf/CVonline/

49
50