Individual difference in behavioral traits consistent with time and context can contribute to the formation and maintenance of regional sea turtle populations, inducing preferences in migratory routes and reductions in the fitness of hybrids. In this case, it is expected that behavioral differences exist among geographically distinct sea turtle populations. However, there is little methodology available for quantifying the behavioral traits of them. This study focused on capturing limb movements from videos to reduce human subjective intervention and the time required for data collection. The captured information of limb movements is compressed along the principal component axes by applying shape analysis. Applying this method to hatchlings of green sea turtles resulted in extraction of limb movements such as rowing and flapping. Categorizing the temporal changes in these movements is expected to enable the quantification of individual differences in behavioral traits.