5 day course
This course aims to develop knowledge of image processing techniques to select appropriate processing techniques for images from a wide range of sources.
This course is suitable for delegates wishing to gain a basic level of competence in the operation of the Erdas Imagine image processing package.
- Introduction: role of image processing methods
- Image restoration and display: digital representation of imagery by scanners, spatial resolution, pixels, spectral resolution, digital format of images.
- RGB and IHS colour space representations
- Corrections for scanner malfunctions
- Image enhancement: contrast stretching (linear, bilinear, gaussian, histogram equalisation and manual), digital filtering in the spatial domain (low-pass, high-pass, high-boost, median and directional)
- Geometric correction: map projections, selection of ground control points, transform equations, resampling methods (nearest neighbour, bilinear interpolation, cubic convolution, sinc x)
- Digital mosaicing
- Classification: density slice, box classification, maximum likelihood and minimum distance algorithms. Supervised and unsupervised training techniques
- Spectral coincident plots and decision tree classification
- Advanced image processing techniques: principal component analysis and formation of eigen images
- Band arithmetic
- De-correlation stretch
- Selection of appropriate image processing system and planning digital processing sequences
- Current classification methods
On completion of this course the delegate will be able to:
- identify a wide range of image processing techniques
- explain the purpose of each process and the underlying mathematical principles
- select appropriate image processing sequences to achieve predetermined objectives
- operate and manage an image processing system
- Integrate image processing techniques into applications of remote sensing