BMVA
British Machine Vision Association and Society for Pattern Recognition
3rd DERA Image Processing and Understanding Course
Programme
27th 29th September 2000
Click here for the registration form
27th September 2000
10:30 Coffee and registration - Nelson Complex, DERA Malvern
11:00 Welcome - Dr Chris West (DERA)
11:10 Introduction: Image Modelling
Dr Margaret Varga (DERA)
- Image basic functions
- Discrete transform
11:40 Image Enhancement
Dr Margaret Varga (DERA)
- Point/corner operations
- Histogram modelling
- Spatial operations
- Transform operations
- Multispectral image enhancement
- Pseudo-colour
12:30 Lunch
13:30 Digital Image Processing and Applications
Professor Mike Brady (University of Oxford)
- What is a "feature"? conventional feature detection, the phase congruency approach: various applications
- Image matching, communication, and learning
- The role of AI: adaptive image analysis systems
14:50 Feature Detection/Extraction/Representation
Dr Paul Ducksbury (DERA)
- Spatial feature extraction
- Edge detection
- Boundary extraction/representation
- Region representation
15:35 Tea and discussion
16:00 Image Registration and Change Detection
Mr Phil Kent (DERA)
- Area correlation
- Feature matching
- Phase techniques
- Re-sampling and Warping
- Change detection
17:00 Breast Image Analysis: a principled approach to image processing
Professor Mike Brady (University of Oxford)
- X-ray image formation: enhancing mammograms
- MRI and contrast-enhanced MRI: breast MRI
18:00 End of Day One
19:00 Evening event (including dinner)
28th September 2000
09:00 Image Database Retrieval
Dr Farzin Mokhtarian (University of Surrey)
- Introduction to image database systems
- Commercial database systems
- Conventional database theory
- Case study of DIMAP system
- Content based retrieval including shape, texture, colour and edges
- Case study of QBIC and SQUID systems
- Video database systems in brief
10:00 Introduction to Wavelet Analysis
Mr Graham Watson (DERA)
- Harr transform
- Orthogonal wavelets
- Bi-orthogonal wavelets
- Variable shape wavelets and wavelet packets
- Feature detection and extraction
11:00 Tea and discussion
11:30 Texture Analysis
Dr Doug Carmichael (DERA)
- Parametric texture models
- Feature based discrimination
- Physical scattering models
- Non-parametric techniques
- Fast classifier algorithms
12:30 Shape Analysis
Dr Farzin Mokhtarian (University of Surrey)
- Concept of shape similarity
- Global shape measures, Compactness, Eccentricity, Rectangularity
- Shape descriptors, Moments, Fourier Descriptors
- Curvature Scale Space Representation
- CSS computation, properties and examples
- CSS matching
- Retrieval under affine transforms
- Application to robust, multi-view object recognition
13:30 Lunch
14:30 Image Projection
Dr Roberto Cipolla (University of Cambridge)
- Orthographic, weak perspective projection, pin-hole camera model and perspective
- Lengths, area, parallelism (vanishing points) and symmetry under affine and perspective projection, co-ordinate transformations, homogenous co-ordinates
- Projection, calibration and pose estimation
- Projective invariants: cross ratio, 2D and 3D invariants
15:30 Image Sequence Analysis
Dr Mark Hodgetts (SIRA)
- Image segmentation using Markov Random Fields
- Basic theory- Optimisation methods
- MRF extension to Image Sequences by exploiting:
- Continuity in Space-Time
- Predictive Kalman Filter Theory
- Introduction to Optical flow
- Horn and Schunks Algorithm
- Recent Advances16:30 Tea and discussion
17:00 Stereo Vision & 3D Reconstruction
Dr Roberto Cipolla (University of Cambridge)
- Stereo vision
- Epipolar geometry
- Essential and Fundamental matrices
- Projective, affine and euclidean reconstruction
18:00 End of Day Two
19:00 for 19:30 - Course dinner
29th September 2000
09:00 Image Compression
Dr Margaret Varga (DERA)
- Lossless and lossy compression techniques
- Still image compression
- Video image compression
- Compression standards
09:45 Colour Processing
Professor Maria Petrou (University of Surrey)
- Colour spaces
- Colour texture
- Colour constancy
- Colour models
10:45 Tea and discussion
11:15 Supervised and Unsupervised Classification
Dr Andrew Webb (DERA)
- Supervised learning
- Non-supervised learning/clustering
12:15 Lunch
13:30 Overview of Fusion of Information
Professor Maria Petrou (University of Surrey)
- Bayesian systems
- Fuzzy Logic
14:30 Real-time Hardware Architectural Design
Dr Bill Smith (DERA)
- Requirements for real-time image processing architectures
- Multiprocessor architectures
- Interprocessor communications
- Processor technologies GPP, DSP and FPGA
- Software frameworks for real-time systems
15:30 Tea and end of course