Isolating Parts from Surface Data

Gocator lets you isolate and then measure parts in two different ways: by configuring the Part Detection panel on the Scan page in the web interface (for more information, see Part Detection); and using one of two Surface measurement tools (for more information on these tools, see Blob and Segmentation).

The following table lists several differences between the two methods. A key difference however is that part detection extracts scan data that is identified as a "part" and outputs it as a separate frame. This lets you use any measurement tool on parts individually. Note however that parts must be clearly separated and be relatively consistently spaced for the part detection algorithm to separate the parts. In general, if you can successfully isolate parts using part detection, use this method rather than the Surface tools.

With the two Surface measurement tools on the other hand, areas are not extracted as individual frames, and for this reason you can't easily apply measurement tools to the areas individually: given that damaged areas may appear anywhere in the source surface data, you can't know where to place the measurement tools. The individual parts are however available for consumption by an SDK application or a GDK tool. (For information on the SDK and GDK, see Development Kits.) The main advantage of these tools is that they can separate objects that are touching. Although you can't apply other measurement tools to the identified blobs, the tools do provide measurements such as length, width, and area, which lets you handle common pass/fail needs.

Main Differences Between Part Detection, Surface Blob, and Surface Segmentation

	Part Detection	Surface Blob	Surface Segmentation
Allows output of individual surfaces to separate frames	Yes	No	No
Allows separating touching objects	No	Yes - Limited Through Open filter, some connections between parts can be separated, but the control is more limited than with Surface Segmentation.	Yes
Supports background present	Yes Height threshold must be set above/below background	Yes Height threshold must be set above/below background	Yes Full support in firmware v6.0 and later
Supports background with significant tilt or intensity gradient	No Fixed height threshold is used	No Fixed height threshold is used	Yes Adaptive threshold is used
Integrated Width/Length/Area measurements	N/A	Yes	Yes
Includes circularity and convexity filtering	No	Yes	No
Fast operation	Yes	Yes	No
Finds objects above or below background	Yes	Yes	Yes But requires careful region placement