VMS has the capability to run a different blob algorithm through the blob finder tool that allows for more advanced attributes of the blobs to be calculated and analyzed. Enhanced Blob is only available with the Advanced Image Processing Package (Image Processing, Enhanced Blob, and Image Stitching). Contact email@example.com for a quote and trial license.
Enhanced blob is useful for providing extra ways of analyzing, categorizing, and detecting the presence of non-geometric features like defects. This capability is only available through the advanced view of the blob finder as shown below.
The fill level value in the blob finder exclusively used with enhanced blob. The enhanced blob algorithm can find so called “child blobs” which are blobs within a blob. A fill level of one will find a blob within a blob. A fill level of two can find a blob within a blob within a blob. And so on with larger values for fill level.
The biggest advantage of enhanced blob is that the blob points are in order around the periphery of the blob. For regular blob, they are in raster order along each horizontal line. This allows better analysis techniques to be applied since the points can be analyzed in a more meaningful way and do things such as:
- Blob combination: Through the merge_blobs() function two sets of blobs can be analyzed and if any blobs are overlapping they can be combined into a single blob instead of two. Note, the two sets of blobs can come from different fields of view.
- Check for intersecting blobs: Through the are_two_blobs_crossing() blobs function, it can be checked if two blobs intersect each other. These blobs can also come from two different fields of view.
- Complex blob attributes: Through the calc_blob_feats() and calc_blobs_feat() functions various attributes can be calculated in addition to the regular attributes in the blob structure. The most commonly used ones are below, along with an illustration.
- Angle of the blob along the principle axis
- Perimeter of the blob
- Compactness of the blob. This is used to determine how close to a circle the blob is. The closer this value is to 1.0, the closer it is to being a circle.
- Length and width of the blob along the principle axis.
- Left, right, top, and bottom extents of the blob.
These blob attributes provide a lot of valuable information, along with the basic blob attributes, such as area and position to effectively identify, categorize, and analyze non-geometric shapes in VMS.
Stay subscribed for future articles detailing usage of merge_blobs() as mentioned above as well as image masking using blobs. Contact our software and applications department with questions and for support.