Comparative analysis of segmentation methods for straw counting using LabVIEW

Counting of semen straw is challenging task due to a smaller size of straws, non-uniform bunch, and presence of liquid nitrogen. Straws are placed in the compact bunch with liquid nitrogen surrounding for preservation. Straw diameter is tiny (0.2 mm). Therefore, manual counting is tedious and produces inaccurate counting results. In such condition, automatic counting can be achieved by using image processing. Straw counting process consists of three basic steps that are image acquisition, image segmentation, and counting. This paper proposes a comparative analysis of segmentation methods used in straw counting. Area of the bunch of straws is segmented by using thresholding, morphological erosion operation and advanced morphology like 'remove small object'. The particle filter is used for selection of area required for one straw. All operations performed by using LabVIEW software.