Instead of using pixel images as regularized correlation searches do, features of an image (geometric shape data) are extracted to specify the location.
For example, a geometric shape search is better than a regularized correlation search in the following cases:
Because only geometric configuration information is included, data is more manageable than before.
Geometric configuration matching results show the XY coordinates and scores (NOT correlation factors). Generally, geometric shape pattern matching is said to have higher repeat positional accuracy than a regularized correlation search.
GradFinder®conducts geometric configuration pattern matching for high-speed, high-accuracy image detection with a proprietary ViSCO algorithm.
The pattern matching function embedded in the alignment tool extracts geometrical features (e.g., edges) from the image. The position with the highest matching level of geometric configuration features is used as the answer.
Our proprietary feature extraction technology, adoption of matching methodology to reduce blur/noise influence, and consideration of various lighting conditions/variation lead to the realization of product stability and improved reliability.
■Alignment, as guidance for robots or the like
■Alignment, as preprocessing that inspects or measures
The application range of geometric configuration pattern matching is the same as that of regularized correlation searches, but geometric configuration pattern matching offers more options because it can be used to deal with changes.
We are an expert group in the field of visual inspections and image processing inspections. We are not just a manufacturer. Instead, we are a visual inspection and image processing inspection system manufacturer that combines knowledge of and experience related to image processing algorithms, optical technology, electronics, and machinery in order to provide comprehensive consulting as a development engineering company.