Detect circles

POST /api/image-detect-circles

Detect circles in an image. Returns each circle's centre and radius. Useful for counting coins / pills / ball bearings, and finding wheel-like features in CV pipelines.

Finds circles in an image using the Hough Circle Transform. Returns each circle's centre `(x, y)` and `radius` in pixels. Tunable: minimum / maximum radius (in pixels), distance between circle centres, accumulator threshold. Defaults work well for everyday-scale objects (coins, pills, ball bearings).

Inputs

Name	Type	Default	Description
image*	file	—	Input image bytes.
dp	number (1…4)	1	Inverse accumulator resolution ratio.
minDist	number	20	Minimum centre distance between circles in pixels.
param1	number (1…500)	100	Upper Canny threshold used internally by Hough.
param2	number (1…500)	30	Accumulator threshold — lower means more (and noisier) circles.
minRadius	number	0	Minimum circle radius (px). 0 = unbounded.
maxRadius	number	0	Maximum circle radius (px). 0 = unbounded.

Response

Modes: json. Cache: yes (24h TTL).