Image processing algorithm for automatically measuring morphological features of spirulina filaments in microscopy images

The morphological features of Spirulina can be related to environmental conditions and can affect biomass harvesting; thus, monitoring these parameters is desirable to assess the state of the culture and to establish favourable conditions for achieving easy-to-harvest filaments, improving biomass productivity. However, obtaining significant measurements of these morphological features can be slow and laborious since it is normally manually. This work presents a new methodology for automatically measuring the morphological features of Spirulina filaments on microscopy images. A novel algorithm is presented that uses the variance of the distances between pixels on each half of the contour to identify filament ends. Unlike previous methods, this approach does not assume that filaments are coiled around a straight axis. Once the endpoints are identified, the filaments' morphological features (length, width, helix angle, and diameter) can be estimated through the fitting of lines and Bezier curves, as explained in sections 3.4–3.6. The addition of editable objects to represent the fitted parameters is presented, allowing for manual adjustment of the results. To evaluate the proposed methodology, the morphological parameters of some Spirulina filaments were estimated and compared to manual measurements. The results of the proposed algorithm had close agreement with the manual measurements with variations below 4% in filament length; a noticeable result considering that this feature is the most time-consuming to measure manually. Overall, the results indicated that the proposed methodology can automatically and reliably estimate the morphological parameters of Spirulina filaments.