Show me the way

I feel like a mountaineer standing at base camp and looking up at the peak. There’s an impressive array of equipment beside me, but I can’t figure out what I need for the climb or which route to take.

That’s how I feel right now with the IMPACT machine vision software package from PPT. I can click through an impressive range of tools, including some that to my knowledge are unique to PPT, such as their utilities for joining images together, but I can’t quite figure out how to bend them to my will. I have a task to perform - OCR on a printed part – and I have the bones of an algorithm in mind, but I’m lacking a guide. What I need is a Sherpa to lead the way.

I mean that in a metaphorical sense of course. A lean, weatherbeaten mountain guide in his Yak-skin coat would look a little odd sat at my desk. What I mean by guide is a tutorial that takes me through the process of developing an application. Unfortunately, IMPACT appears to lack such an add-on, which is a great shame.

I’m becoming ever more impressed as I explore IMPACT. There are a lot of tools, and they seem to function effectively, What’s more, they link together with a kind of intelligence, saving me the work of defining origins and simplifying the business of repositioning tools to cope with parts that move from image to image. So far, so good. But I do feel I’m climbing not so much a learning curve as a learning cliff. The manual is frankly of limited help: by describing how the individual tools operate you could say it’s giving me the hand-holds I need to climb the rock wall, but when it comes to figuring out the route to take, I’m on my own.

In short, I think IMPACT will turn out to be a very capable package, but so far it’s looking rather hard to learn. I will stick at it for a few more hours, hoping things will drop into place, and I’ll put out another post on the topic soon.