Basic Information

  • Course Code: GNR607
  • Course Name: Principles of Satellite Image Processing
  • Course Offered In: 2023-‘24
  • Semester Season: Autumn
  • Instructors: Bal Krishan Mohan
  • Prerequisites: None
  • Difficulty (1 being easy and 5 being tough): 3

Course Content

Introduction to Digital Image Processing Sensors for Remote Sensing Histogram and Image Enhancement Methods Neighborhood Operations Various Filters (e.g., Sobel Filter, Prewitt) Hough Transform Color Models Principal Component Transformation (PCT) Data Fusion and Band Arithmetic Texture Analysis Principles Fourier Transform and Filters Morphological Image Processing Image Classification Methods Supervised vs. Unsupervised Classifiers and Subtypes Likelihood Calculation NDVI Calculations Various small topics related to remote sensing

Feedback on Lectures

This course is an introductory one that covers a lot of topics, but not in much depth. The professor uses both the board and slides for teaching, and you can get by with just the slides if you put in some effort to understand the material. The lecture pace is generally slow, leading to 2-3 additional classes(sometimes 2-3 hrs) during the semester. The professor’s delivery is slow and lacks enthusiasm, but if you attend the lectures, a one-hour review on the same day should be enough. If you have a background in Electrical Engineering, especially with Fourier transforms (like in Signals EE229), you’ll find that part of the course particularly useful.

Attendance is compulsory with regular checks. The professor mentioned that falling below 75% attendance could result in a DX grade, though no one ended up getting it.

There is a project towards the end of the course with multiple topics to choose from. It’s relatively easy, especially with tools like GPT, and requires a small PowerPoint presentation for submission.

Feedback on Evaluations

2 Quizzes: 10% each Midterm Exam: 20% Final Exam: 40% Project: 20% Reading the slides and understanding the concepts is more than enough to do well. However, the course includes some factual knowledge about remote sensing that can be hard to memorize, and the professor does include 5-6% of questions on these facts in quizzes and the midterm.

Grading was a bit disappointing the year I took the course, although the same professor graded very generously the previous year (Autumn 2022).

Study Material and Resources

For a deeper dive into the topics, the professor recommends “Digital Image Processing” by Rafael C. Gonzalez and Richard E. Woods.

Follow-up Courses

GNR602 - Advanced Methods in Satellite Image Processing

Final Takeaway

This course is a perfect introduction to satellite image processing, covering a broad range of topics. If you’re curious about remote sensing or image processing, this is a great starting point.