基于Web端舰载机手势识别仿真系统设计

Design of the carrier aircraft's gesture recognition and simulation system based on Web terminal

主要应用于航母舰载机起飞及着舰的地面飞行模拟仿真,仿真系统基于深度学习的卷积神经网络方法开发。根据采集的图片数据集训练出用于手势识别的多层卷积神经网络模型,再经过大量的数据集训练与评估获得分类数据并实现手势识别;使用python编程语言搭建Web服务器,识别后的分类数据用来驱动Web端的三维模型,并采用Ajax技术实现分类数据在Web端的实时自动刷新。仿真结果表明本系统能够较精确识别航母手势,达到了预计目的。

The system is mainly used in carrier-based aircraft＇s take-off and landing flight simulation of the ground, which is a convolution neural network method based on the deep learning. According to the collected image data set, the multi- layer convolution neural network model is trained for gesture recognition, and then the data set is trained and evaluated to obtain the classification data and achieve the recognition of the gesture. ;Using python programming language to build Web server, the classi- fication results after recognition is used to drive the 3D model of the Web, and the use of Ajax technology is to achieve the classifi- cation data＇s automatic refresh. Simulation results show that the system can accurately identify the aircraft carrier gesture, and achieve the expected goal.