Ask the Experts

There are a number of ways, some of which are more difficult to achieve than others. Use lighting to light up the different heights differently. Use 3D technology to give you a height difference if it exists. Image formation with regards to machine vision is about getting the image to show you the thing that you are looking for. Its much easier to answer whilst looking at samples. If the samples are opaque, but are very similiar if the one at the front has a height difference, then 3D machine vision may be your answer. Good Luck.

I personally would first look at the material of the part versus the material of your background. I.E. If you are inspecting a metal part and your background is rubber, my initial thought would be an infrared lighting solution. The rubber will absorb the infrared light differently then the metal part giving you a starting point to create contrast. If that alone does not create enough contrast, I would continue testing with filters. If your part is made from a similar material as your background, darkfield illumination would be something to consider as this should allow you to project majority of the light onto the surface of your part. This should illuminate the surface your part while leaving your background dark.

I would suggest using the 3DPIXA from Chromasens, GmbH: https://www.chromasens.de/en/products/line-scan-cameras-3d-inspection It provides high accuracy 2D and 3D measurement with one device. Using it, you can extract the height information to locate the object and perform 2D and 3D analysis with the image and height data, making it easy to find the object with high precision. The 3DPIXAs are available in different configurations and resolutions. I can help you find the right one.

Dear Radhakrishnan. We have with solino a technology in place to analyze the reflections of surfaces with an RTI aproach and a photometric stereo Input. See here https://www.opto.de/en/software/solino/ That could do the job. Feel free study it it something for you. Best regards.

There are several options, depending on the part and background, if it's the same color - may be it's a different Material, then you can use different wavelength where you can separate it - in different wavelengths, similar colored parts for the human eyes look different - e.g. in NIR IR or UV or any other spectral, if the part has a dimension you could also use a 3D camera to get it - if there is a height you get the part by the height. If you need further information, please write to [email protected] and send some images so that I can give you a precise answer. Best regards Michael

Hello, with an application like yours you may want to implement a contrast invariant technology. Usually this can be done by using a 3D technology where you can detect features regardless of surface color. LMI Technologies has a variety of products that can help with applications like yours. Please let us know if you still need to find a solution to your application. Thank you.

There are some traditional computer vision techniques like edge detection that may prove useful. If that doesn't work you may want to try a 3D line scan or area scan camera.