Morphogenesis of the silica based cell walls of diatoms, a large group of microalgae, is a paradigm for the self-assembly of complex 3D nano- and microscale patterned inorganic materials. In recent years, loss-of-function studies using genetic manipulation were successfully applied for the identification of genes that guide silica morphogenesis in diatoms. These studies revealed that the loss of one gene can affect multiple morphological parameters, and the morphological changes can be rather subtle being blurred by natural variations in morphology even within the same clone. Both factors severely hamper the identification of morphological mutants using subjective by-eye inspection of electron micrographs. Here we have developed automated image analysis for objectively quantifying the morphology of ridge networks and pore densities from numerous electron micrographs of diatom biosilica. This study demonstrated differences in ridge network morphology and pore density in diatoms growing on ammonium rather than nitrate as the sole nitrogen source. Furthermore, it revealed shortcomings in previous by-eye evaluation of the biosilica phenotype of the silicanin-1 knockout mutant. We anticipate that the computational methods established in the present work will be invaluable for unraveling genotype–phenotype correlations in diatom biosilica morphogenesis.