Research On Gait Recognition Algorithm In Sensor Passway

In the intellectualized century--21 century, mode recognition technology which was based on artificial intelligence technology have had a better progress. As security and sensitivity places (e.g. bank, airdrome……) require the intelligence control system, human identification at a distance is being paid attention. In the mode recognition area, it is a great challenge to use gait recognition technology to distinguish different people theoretically and practically. So many investigators are interested in this area. A simple approach of gait recognition based on motion patterns was presented in the present paper. It is applied in sensor passage primarily, identifying walkers and objects.Biological character recognition technology was explained and some familiar technology about it was introduced simply in this paper. Gait recognition technology was presented in detail and gate composed of sensor passage was studied as well.It has been proved that the signals which come from his/her ankle when a person walks show a special braided pattern(XYT pattern). Analytically, it is found that in the sensor passage, since the sensor is kept out when one walks, a series of discrete spots come from the covered sensors. These spots which are defined as twisted-pair, also show a special braided patterns. According to such a special biological feature, this paper identifies the walkers and objects in the passage, using a simple and effective algorithm with four kinds of identifying modes. We suppose three sensors located in the passage. The identifying algorithm records the beginning time and the whole time the three sensors is covered. Via the calculation and comparison of the time recorded, we estimate whether the twisted-paired pattern is brought. If it is present, person goes through, else object goes through.