This paper presents the image analysis algorithms developed for a microscopic system that specifically evaluates cotton properties from longitudinal views. The system utilizes a motorized stage to transport a sample slide to grab sequential images of fiber segments, transversely scans the detected fibers in the images, and outputs both fineness and maturity measurements of cotton based on the statistic data of the massive scans across fibers. The experiments showed that the longitudinal data were repeatable and consistent with the data obtained from other testing methods.