Design of embedded dance teaching control system based on FPGA and motion recognition processing

摘要

The complex human activities, to the dance teaching, it has not been constrained, it is a difficult task for computer vision. Field-programmable gate array (FPGA) a classification of the movement of the Indian classical dance by using the powerful features of the built-in system tools. Human behavior recognition for this work, classical dance education in India from offline and live performance to the online control of material in the carried out of the guidance record of audio and dance, in. Another theme and 10 familiar dance of 200 m / bills have created, is the offline data of guidance form of a variety of classical dance in various postures. The data collection of a variety of topics. Or education attitude accounted for 60, Dance education in both cases. FPGA training and eight different sample size, will be performed by each of the multi-group object. The remaining two samples are tested, it has been trained on the FPGA. Different FPGA architecture design, and excellent recognition accuracy, because to get our test data. Compared to the same data set and of other classification model report, the recognition rate.