In this paper, we analyze public transport passenger movement data to detect typical patterns. The initial data consists of smart card transactions made upon entering public transport, collected over the course of two weeks in Saint Petersburg, a city with a population of 5 million. As a result of the study, we detected 5 classes of typical passenger movement between home and work, with the scale of one day. Each class, in turn, was clusterized in accordance with the temporal habits of passengers. Heat maps were used to demonstrate clusterization results. The results obtained in the paper can be used to optimize the transport network of the city being studied, and the approach itself, based on clusterization algorithms and using heat maps to visualize the results, can be applied to analyze public transport passenger movement in other cities.