The cornea is the most sensitive biological membrane in the human eye which consists of five layers viz. epithelial, Bowman, stroma and Descemet's and endothelial membrane. Endothelial layer in the cornea is the most vulnerable membrane among all the five membranes. This layer is significant as it allows free flow of aqueous humor fluid inside and outside of the cornea. Fuch's dystrophy (FD), Advanced Fuch's dystrophy (AFD), Posterior Polymorphous Corneal Dystrophy (PPCD), and Iridocorneal Dystrophy (ICD) are the major dystrophy's that affects the endothelium layer directly. The cells are drastically decreased when the dystrophies affect the endothelium layer. Morphological operations (MO1 & MO2) are used to estimate the endothelium cell density. A novel hybrid algorithm (HA) using Confocal Microscopy (CM) images have been proposed to identify the mean value of unique cells, standard deviation, area, cell densities and its quantitative values for the endothelial layer. The median filter is used to eliminate the noise. The estimated MSE (Mean square error) and PSNR (Peak Signal to Noise ratio) represented the filtered image for further processing. Altogether Eight (8) confocal endothelial dystrophic and two (2) normal images were used for analysis. The results harvested with the HA are comparable with standard manual (SM) and conventional cell density approach. The average error differences of MO1 & MO2 and S-PSO (Snake Particle Swarm Optimization) to the HA is 12.14%, 20.46%, 9.73%, and 8.00% respectively. The average inspection time of HA is 61.48 ms with standard deviation (SD) of 1.59 ms. The proposed algorithm showed high accuracy with a very low processing time of 58.57 ms which is suitable for detection of the membrane at an early stage to diagnose any disease related to the endothelial layer. © School of Engineering, Taylor's University.