Blob Detection and Labelling

Donald G. Bailey · 2011

The processing required for blob detection and labelling falls within the intermediate processing level within the image processing pyramid. The input is all in the form of image data, with the output in terms of segmented regions or even data extracted from those regions. While there are many possible algorithms for all of these operations, the primary focus in this chapter has been on stream processing. Full connected components labelling is required to take into consideration the connectivity. Chain coding is an alternative region labelling approach, which defines regions by their boundaries, rather than explicitly associating pixels with regions. While other intermediate and high level image processing operations can also be implemented in hardware on a field programmable gate array (FPGA), these operations are selected as representative operations where significant benefits can be gained by exploiting parallelism. Controlled Vocabulary Terms field programmable gate arrays

Read the paper · More papers on PaperTik