Abstract: Fluorescence microscopy imaging is one of the important methods for the study of the dendritic spine structures. Most of the current morphologic analyses in low SNR images involve a significant component of manual labor and is susceptible to operator bias. An algorithm based on endpoint match-searching and segmentation-sampling curve fitting was presented to automatically detect the edge of the dendrite, separate dendritic spines and calculate their size, number and density in complex images. Analysis of the results shows that it provides an efficient, accurate tool for dendritic spine analysis.

Key words: dendrite, spine, segmentation-sampling, match-searching