Knowledge · Code reading
A selection router for engineers: when a smart code reader beats a PC-based setup, why Direct Part Marking needs lighting first, and how to move from samples to a production-ready Computer Vision ID station.
Geometry of decisions
Code reading fails for predictable reasons: contrast, curvature, motion, working distance and how the mark was applied. Fix the optical path before you debate firmware features.
| Decision | Choose this when | Watch for |
|---|---|---|
| Smart code reader | Dedicated ID/OCR station, stable FOV, I/O to PLC, limited vision software scope | Mark quality, lighting and presentation still dominate |
| PC-based camera + software | ID is one tool among inspection, measurement or robot guidance on the same image | Integration effort and deterministic timing |
| DPM on metal / curved parts | Dot peen, laser or inkjet marks with low contrast or shine | Lighting geometry first; feasibility needs real samples |
| Line-scan ID | Continuous webs, long parts or high-speed conveyors | Encoder sync and illumination uniformity |
Routes
Pick the path that matches your question. Commercial feasibility lives on the solution page; definitions and hardware shortlists live here.
Application framing, when Sedeco configures ID stations, and how to start a feasibility case with samples.
What Direct Part Marking is, why contrast and curvature hurt reads, and why lighting comes before decoder settings.
How industrial readers work versus an area scan camera with software, and what drives stable reads.
Hikrobot MV-ID families: selection criteria and links into the catalog for fixed, line-scan and handheld devices.
Lighting fundamentals that decide whether a Data Matrix on metal is readable at all.
Anonymized application story on tough codes and what the optical setup had to solve.
Send parts or high-quality images of the codes you need to read. We assess lighting, optics and reader family against the real mark, not a brochure claim.