The invention relates to a non-contact
physical fitness testing
system and testing method based on a depth image. The
system is formed by a
physical fitness testing control module, a sit-up testing module, a push-up testing module, a voice recognition module, and a test feedback module. The depth image is obtained through using a Kinect somatosensory interactive device, a
finite state machine framework is used to count the motion of a tested
human body, a voice
recognition algorithm is used to automatically start or end a
physical fitness test, and a related test result is fed to a user at the same time. According to the
system and the method, the capture and recognition of the depth image are fused, the limb information of the user is obtained under the premise that the limb of the user does not carry any apparatus, and the precise exercise test counting is carried out. In addition, the system can
record automatically according to the matching of a user voice
signal, other human intervention is not needed in a detection process, at the same time the related test result can be fed to the user according to detection data, the user can achieve the maximum physical potential of the user, the exercise efficiency and effect are improved, and a self monitoring platform is provided for non-professional sports players.