Lab 1: Introduction to Digital Image Processing
Objective: Familiarize students with the basic concepts of digital image analysis and explore the
tools used in image processing.
Activities:
Overview of digital image structure: pixels, resolution, and bands.
Introduction to software (e.g., QGIS, ENVI, ERDAS IMAGINE, or ArcGIS).
Loading satellite imagery and exploring different image types landsat (optical,
radar, and thermal).
Visualization of single and multi-band images (true color, false color).
Task: Visualize different satellite images and interpret their content.
Lab 2: Image Preprocessing
Objective: Learn the importance of preprocessing steps like radiometric and geometric
corrections.
Activities:
Perform radiometric correction: atmospheric correction using methods like Dark
Object Subtraction (DOS).
Perform geometric correction: rectifying an image using ground control points
(GCPs).
Resampling techniques: nearest neighbor, bilinear, and cubic convolution.
Task: Apply radiometric and geometric corrections to a Landsat or Sentinel-2
image and assess the quality.
Lab 3: Image Enhancement Techniques
Objective: Enhance satellite images to improve interpretability for specific applications.
Activities:
Histogram equalization and stretching (contrast enhancements).
Spatial filtering techniques (smoothing, sharpening).
Edge enhancement using Sobel and Laplacian filters.
Task: Enhance a raw image using contrast stretching and filter techniques, and
compare the results.
Lab 4: Spectral Indices Computation
Objective: Use spectral indices to analyze specific features in satellite images.
Activities:
Calculation of NDVI (Normalized Difference Vegetation Index) for vegetation
monitoring.
Computing other indices: NDBI (Built-up Index), NDWI (Water Index), and SAVI (Soil-
Adjusted Vegetation Index).
Task: Compute various indices for an image and analyze different land cover features
(vegetation, water, urban areas).
Lab 5: Image Classification Techniques
Objective: Introduce students to supervised and unsupervised image classification techniques.
Activities:
Supervised classification: Define training areas and apply Maximum Likelihood
Classification (MLC).
Task: Perform both unsupervised and supervised classification on the same image and
compare the results.
Unsupervised classification: Perform K-means or ISODATA clustering to classify an
image.
Lab 6: Accuracy Assessment of Image Classification
Objective: Assess the accuracy of image classifications using confusion matrices and other
metrics.
Activities:
Introduction to accuracy assessment: confusion matrix, overall accuracy, user’s
accuracy, producer’s accuracy, and kappa coefficient.
Ground truth data collection and comparison.
Task: Conduct accuracy assessment of the supervised classification performed in Lab 5,
using reference points or ground truth data.
Lab 7: Object-Based Image Analysis (OBIA)
Objective: Use object-based methods for image classification to complement pixel-based
approaches.
Activities:
Image segmentation: segment the image into meaningful objects (multi-resolution
segmentation).
Classifying segmented objects based on spectral, shape, and texture information.
Task: Segment an image and classify the objects using an OBIA approach. Compare with
pixel-based classification.
Lab 8: Change Detection Analysis
Objective: Identify and quantify land cover changes over time using multi-temporal satellite
images.
Activities:
Preprocessing multi-temporal images (e.g., aligning imagery from different years).
Change detection techniques: image differencing, post-classification comparison, and
NDVI differencing.
Task: Perform change detection analysis for an area using Landsat or Sentinel-2 data over
two different years.
Lab 9: Feature Extraction and Image Fusion
Objective: Extract features such as roads, buildings, and vegetation, and explore the concept of
image fusion.
Activities:
Feature extraction techniques using edge detection, texture analysis, and spectral
properties.
Image fusion: combining different types of imagery (e.g., optical and radar) for
improved analysis.
Task: Extract features from an image and perform image fusion to combine optical and
radar data.
Lab 10: Accuracy Assessment and Validation for Land Use/Land Cover (LULC)
Maps
Objective: Focus on advanced accuracy assessment techniques for LULC classification.
Activities:
Introduction to reference data sources: field data, high-resolution imagery, Google Earth.
Performing accuracy assessment using different reference data types.
Comparison of results with various reference data.
Task: Perform a comprehensive accuracy assessment using different sources of reference
data, as per your research interests.
Final Project:
Objective: Apply the learned techniques to a real-world scenario.
Activities:
Select a study area and acquire suitable satellite imagery.
Perform preprocessing, classification, and accuracy assessment for land use/land cover
mapping or another application of your choice.
Submit a lab report including all steps from image preprocessing to final classification
and accuracy assessment.