Research on tennis-assisted teaching assessment technology based on improved dense trajectory algorithm

Abstract

With the continuous development of artificial intelligence technology, tennis robots begin to enter people's lives. This study proposes a tennis-assisted teaching evaluation method based on improved dense trajectory algorithm. The results show that the best recognition effect is obtained under the division method when the division parameters nσ and nΓ of non-fixed-length trajectories are 2 and 3 respectively. The calculation effect obtained by Chebyshev distance formula is better, and the distance between various types of actions is mainly distributed around 0.2, and the actual distance existing between them and the standard action is smaller. The BD rate of 4K video increases from 11.45% to 32.98%, while that of 8K video increases from 1.82% to 7.61%. The more the number of tile divided by the system, the more the performance is lost, and globally, the coding performance affected by the optimised MCTS is acceptable. From the global view of the test, it is still possible to accept the transmission delay caused by the faster code stream fusion to the system.