Parts inspection based on Image.

[MogaRaghu]

There is a requirement for final acceptance inspection in a assembly line. Right now it is based on visual inspection - where the inspector physically looks for completeness of the assembly, fixing of warning labels , part ID label etc. 

I Would like to know what options are available with LabVIEW to accomplish this. In other words there will be a master image file against which a image from a camera (fixed over the assembly conveyor ) will be compared to decide if the part can be accepted. 

Being a first of kind requirement, would like to get some pointers to go about this. 

Thanks !

[wiebe@CARYA]

---

@MogaRaghu wrote:

Being a first of kind requirement, would like to get some pointers to go about this. 

---

Make sure the lightning conditions are good* and stable.

Make sure you have the right* camera setup.

Use NI Vision (or OpenCV through Python) to make the software behave correctly*.

Good, right, correct are all circumstantial. Getting things "good, right and correct" usually comes after years of experience.

Vision can be anything from hard to fairly simple. It won't be as simple as creating an image and comparing it though.

You probably have to dig into template matching, OCR, object detection, object labeling, object measurement in the NI Vision toolkit.

[MogaRaghu]

Thanks for confirming its not going to be easy 😉

Yes it should be a learning and more so since I am new to it.  The Vision ToolKit - thats the hint I required but could not place it immediately.  Will study it and get back .