Turn abandoned patents
into your own filing
into your own filing
This invention describes a software method for breaking down a digital image into distinct parts, like objects or regions. It first roughly divides the image, then lets you define a "model object" from these pieces. Using characteristics of these pieces and the model, a special "fuzzy logic" system calculates how to best cut up the image, for example, by determining how smooth or detailed the edges should be.
Why it matters: While computational power and data availability have vastly increased since 2007, making complex image processing more feasible, deep learning now offers alternative, often more performant, segmentation methods.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI
