Real Image Hough Line Transform

Original Image + Detected Lines Result

Color-coded detected lines | Hover on Hough space → preview line here

Hough Space (θ–ρ Accumulator) Parameter Space

Low votes

Medium

High votes

Peak → line

Hover → preview line | Colored dots = detected peaks

Edge Map (Sobel) Click → sinusoid

Click a bright pixel → its sinusoidal curve appears below (max 5)

Sinusoidal Curves of Selected Pixels Intersection Analysis

← Click pixels in the Edge Map to see their sinusoids here

💡 Hough Transform Pipeline on a Real Image

Step 1. Sobel filter → compute gradient magnitude per pixel → threshold to detect edges
Step 2. For each edge pixel (x,y), sweep θ = 0°–180° and compute ρ = x·cos(θ) + y·sin(θ)
Step 3. +1 vote in the accumulator cell at (θ, ρ)
Step 4. Accumulator peaks = lines shared by many edge pixels → deduplicated with NMS

Click any pixel in the Edge Map to see its sinusoidal curve in the bottom-right panel.
Curves from collinear pixels all intersect at one point in Hough space.

📊 Processing Stats

Image size—

Edge pixels—

Voted pixels (sample)—

Max votes—

Lines detected—

Clicked pixels0