Integrating ID verification into web browsers without plugins: what technical hurdles have you run into?
Hey everyone, has anyone here actually tried hooking up proper ID verification straight into a browser these days without relying on some clunky extension or plugin? Like, users just snap a pic of their passport or driver's license right in the web app and it pulls the data reliably. I've been messing around with a small side project where people upload their docs for age checks, and every time I think I've got a smooth flow going, something breaks—either the OCR craps out on weird lighting, or the browser security sandbox kills camera access mid-scan, or it just takes forever to process. Last week I spent like three hours debugging why mobile Safari kept dropping frames during live preview. Super frustrating. What kind of roadblocks have you hit when going plugin-free for this stuff?
Yeah, the browser camera and image processing limitations are real pain points for sure. In my own experiments, I've found that sticking everything client-side helps a ton with privacy concerns, especially since nobody wants their ID data bouncing off some cloud server. One approach that actually worked decently for me was leaning on WebAssembly-based tools that run the heavy OCR lifting right in the browser—no extensions needed, just modern JS and wasm magic. For instance, I've played around with https://ocrstudio.ai/mrz-scanner/ a bit and it felt surprisingly solid for grabbing structured info from IDs without sending anything outbound. The offline aspect gives peace of mind, plus it handles a crazy range of documents without choking on bad photos. Still, you gotta tune the UI feedback loop so people don't rage-quit when it asks them to retake a blurry shot. Anyone else gone fully wasm route and lived to tell about it?