Camera

[Devender]

Hi all, For Ware house management, I need to inspect 72 X 25 area. Can any one suggest which is the best Camera OR any other sensor to inspect large areas. Thanks and Regards Devender T

[ScottSmith]

    It depends.  72 x 25 is in what units?  Feet, centimeters, meters, microns?  😉   And by "inspect", it depends again how many different kinds of items, of what sizes, under what lighting/contrast, via what method(s), you'll be inspecting.  E.g. seeking presence/absence of one item, a "blob" against a flat field, is easy.   With multiple items, differential lighting and shapes, and depending on if recognizing by shapes only or with bar codes, whether camera will be fixed or movable, lens selection, ability to control lighting (or not), camera/sensor selection is just one component of the solution, but can't precede answers to some of those questions.

Regards,

Scott

[Devender]

Sory for not providing recourses my need is I have to identify the objects of variable size( approximately one meter length and few centimeters diameter ). The size of the inspecting area is 72 X 25 square meters. I would like to know how many cameras are required and what type of lenses I have to use. Thanks and Regards Devender

[Jaidev]

Hi Devender,

Your target object seems very small compared to the area under inspection. If you really want to scan an area that 72x25 sq. meters, you would need a camera mounted very high with a very nice lens. If you just want to inspect rods (or any long-ish object), it might be a good idea to have a conveyer belt that moves the object under a fixed camera.

Jaidev

[BruceAmmons]

To cover the entire field with one camera, you would need a bare minimum of 7500x2500 resolution, which gives you one pixel per centimeter.  Depending on the part, you may need even more resolution.

One possibility is breaking the area into smaller regions and using one camera per region.  You would want them to overlap a little bit.  Depending on the resolution of each camera, you might have 3x1, 6x2, 8x3 cameras.  This starts getting pretty expensive.

Another possibility is a single camera mounted on a tilt/pan mechanism.  You could point the camera to each region, snap a picture, then move to the next region.  This would be much cheaper, but your viewing area has to remain static (not moving) and you have to be able to search each area independently of the others.  It isn't as fast, either, so you have to be able handle one complete search every few seconds.

If you have plenty of time, you could do a rough search by scanning the whole area, then have a zoom control that zooms in on potential items to verify them.  I have a metrology system that uses this method.  It scans a large area for the items of interest, then zooms in on each item to do more precise measurements.

Other things you will have to consider is contrast between the parts and the background.  If there is not sufficient contrast between them, it will be very difficult to locate the parts.

The more information you can share about your application and its constraints, the more we will be able to help you.

Bruce