6533b859fe1ef96bd12b7717

RESEARCH PRODUCT

Segmentation and feature extraction in capillaroscopic videos

Domenico TegoloCesare ValentiEmiliano Spera

subject

Settore INF/01 - InformaticaComputer sciencebusiness.industryFeature extractionComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISIONWavelet transformVideo sequenceSet (abstract data type)Oral capillaroscopic videos wavelet segmentation feature extractionPeripheral microcirculationComputer visionSegmentationArtificial intelligencebusiness

description

This contribution describes a method to select regions of interest as capillaries of the oral mucosa and to extract their main features useful for real diagnosis purposes. A discrete version of the wavelet transform has been adopted for segmenting the images coming from video sequences acquired by a prototype capillaroscopic, able to put in evidence the red blood flow. A set of proper characteristics is automatically computed for a correct evaluation of the peripheral microcirculation.

yearjournalcountryeditionlanguage
2015-06-25
10.1145/2812428.2812472http://hdl.handle.net/10447/158920