| Week |
Topic |
| 1 |
Introduction, Purpose of the course
? Week Contents,
? RS I flashback,
? Digital image, why need to process images? DP levels (low, mid and high)
? DI history, pixel, image band histogram and features (mean, st dev, type etc.), cumulative histograms, contrast and brightness info from histograms
? RGB colour,
? image resolutions,
? Application with different images and histograms (application areas) |
| 2 |
Digital Image Processing – Pre-processing I (Radiometric correction)
? Purpose of corrections
? Radiometric corrections
o Sensor imbalance (noise(bit error), striping, dropped lines)
o Atmospheric corrections (purpose, haze reduction, is correction necessary? (Yes or No); Absolute atmospheric correction (Empirical Line Calibration); Relative atmospheric correction Dark object subtraction method and Regression method)
o Scene illumination and viewing geometry (Seasonal Compensation, Sun angle correction, Importance for image mosaics) |
| 3 |
Digital Image Processing – Pre-processing II (Geometric correction)
? Why we need (RS images are not maps!), & Purpose!
? Distortions
o Systematic distortions (predictable)
o Non-systematic distortions (random)
o Terrain distortions
? Why is geometry important?
? When to rectify? (Mosaicking, scene to scene comparison etc…)
? Image to map (rectification)
? Image to image (registration)
? Hybrid method
? Processing steps
o Locate ground control points
o Compute and test a transformation matrix (Order of transformation)
o Resampling (Nearest neighbour, Bilinear interpolation, Cubic convolution)
? Disadvantages of rectification |
| 4 |
|
| 5 |
|
| 6 |
|
| 7 |
|
| 8 |
Image enhancement – Local operations (High pass filtering, Noise, Signal-to-Noise ratio, Noise suppression) |
| 9 |
Low pass filtering + High pass filtering + Noise removal |
| 10 |
Image transformations - Arithmetic operations (+, -, /, x), multi-band/image manipulation (Rationing, Subtraction, Addition, Multiplication) |
| 11 |
|
| 12 |
|
| 13 |
Image Processing - Classification - Supervised and Unsupervised Classification
PRACTICE + HW 5 – Supervised and Unsupervised Classification |
| 14 |
Classification accuracy (Sample design, Reference data, Error matrix, OA, PA, UA, Kappa
error, issues) |