Segmentation in Bioimaging
Archive ouverte
Edité par CCSD ; John Wiley & Sons, Ltd -
International audience. This chapter surveys basic segmentation techniques and point to tools that make them accessible in practical situations. Segmentation is very often the first step when doing quantification. Other image analysis techniques, such as pixel‐based co‐localisation, optical flow, etc. are not based on segmentation, yet offer useful information condensation that can lead to a proper scientific conclusion. The chapter focuses on the list of integers that comprises an image. Automatic counting of objects is almost always done on images where objects appear bright over a dark background, ideally flat. The set of connected pixels with a shared identity forms an object, from which a contour can be extracted and quantification performed. The chapter focuses on what quantifiable information is required to reach a scientific conclusion, pick an analysing technique that can yield it, find or write a software tool that implements it, and finally acquire images that can be analysed with such a tool.
Consulter en ligne
Suggestions
Du même auteur
