Multi‑Projector Blending System¶

Team G11 · Python / OpenCV · Semester 2025

Project Overview¶

Problem: Two projectors produce overlapping images on a planar screen, creating luminance non‑uniformity (the shared region appears brighter).
Approach: Detect the overlap and apply spatially smooth alpha‑blending masks whose weights vary linearly from 1 to 0 across the overlap band.
Deliverable: A minimal, reproducible still‑image prototype implemented in Python (OpenCV).

Objectives¶

• Define the problem and select a feasible solution path prior to implementation.
• Implement an alpha‑blending pipeline and iterate until seams are not perceptible at normal viewing distance.
• Validate the method with a small set of test images, recording inputs, parameters, and outcomes to ensure reproducibility.

Success criterion: no visible seam or luminance discontinuity in the overlap region.

Scope & Operating Assumptions¶

• Configuration: two projectors and a planar screen; projector optical axes nominally normal (perpendicular) to the screen.
• Media: still images for the initial phase (video may follow).
• Overlap geometry: approximately rectangular; blend weights vary linearly from 1 to 0 across the overlap.
• Calibration philosophy: favor simple, reproducible steps; apply additional calibration only if visible artifacts persist.

Team & Responsibilities¶