xmba15 / simple_parallel_line_fitting

simple parallel line fitting using RANSAC linear least squares

📝 simple parallel line fitting using RANSAC Linear Least Squares approximation

---

• fitting of 2 parallel lines: see details about the formulation in reference [2]

• fitting of 2 parallel parabolas: not the official text-book definition, but here 2 parallel parabolas p₀, p₁ are defined by: ax²+bx+c₀ (p₀); ax²+bx+c₁ (p₁). Then the parallel parabola fitting can be formulated by the following least squares problem:

🎉 TODO

---

• fitting of two parallel lines, demo code
• fitting of two parallel parabola, demo code
• add julia code (just to practice julia)

🎛 Dependencies

---

• python

python -m pip install -r requirements.txt

• julia

pkg> activate .
pkg> instantiate

🏃 How to Run

---

• python

  • demo for fitting of two parallel lines

  python demo_parallel_line_fitting.py

  • demo for fitting of two parabolas

  python demo_parallel_parabola_fitting.py

• julia

  • demo for fitting of two parallel lines

  julia --project=. demo_parallel_line_fitting.jl

  • demo for fitting of two parallel parabolas

  julia --project=. demo_parallel_parabola_fitting.jl

You can also call the demo scripts from julia REPL with include syntax. This will reduce the latency for code loading and compilation for the second run on REPL. However, you will always suffer from loading and compilation latency if running directly from command line.

💎 References

---

• [1] basic theory about least-squares data fitting
• [2] basic theory about parallel line fitting
• [3] a detailed blog about RANSAC 3D line fitting without SVD-based least squares method