Inner diameter measurement of aligned TiO2 nanotubes by computational image analysis

Vertically aligned TiO 2 nanotubes have become a very attractive topic for researchers recently. Their properties can be applied in several industries and are related to the size of nanotubes, for example with their inner diameter; hence measuring this dimension is an important activity. Scanning electron microscopy obtains images of nanotube samples which allows manual measurement of inner diameter through functions of the microscope software. This, however, poses the disadvantage of being rather time consuming, subjective and shows low representative readings. This paper presents a proposal to measure the average inner diameter of vertically aligned TiO 2 nanotubes using image analysis, overcoming the aforementioned drawback. Results showed that image analysis with segmentation by k -means + + clustering can correctly detect a representative number of nanotubes for statistically sound diameter measurement, with a low error value.