人脸与语音识别技术的拖拉机操控系统研究

The Research of Tractor Control System Based on Face and Speech Recognition Technology

针对我国拖拉机智能化水平较低的问题,基于人脸和语音识别技术对拖拉机的操控系统进行了设计。系统主要组成包括中央控制器、人脸识别系统、语音识别系统、行驶控制系统和网络传输系统等部分。为了使拖拉机能够实现人脸和语音识别的功能,对人脸和语音识别算法进行了设计,包括采用KPCA算法、Gabor小波算法和FLD判别法进行人脸特征提取,并采用HMM算法进行语音识别。为了验证人脸和语音识别系统的性能,分别对其进行试验,结果表明:人脸和语音识别系统的识别准确率分别达到了96%和93%,可以满足对拖拉机的智能控制。

Aiming at the problem of low level of intelligent tractor in China, the control system of the tractor based on face and speech recognition technology was designed. The system was constituted of central controller, face recognition system, speech recognition system, driving control system and network transmission system. In order to meet the function of face and speech recognition, the algorithm of face and speech recognition was designed. The KPCA algorithm, Gabor algorithm and FLD discriminant method were used to extract the face features;the HMM algorithm was used to recognized speech. To verify the performance of the face and speech recognition system, the tests were carried out respectively. The test results show that the accuracy of the face and speech recognition system could respectively reach 96% and 93%. It could meet the intelligent control of the tractor.