| The purpose of sign language recognition research is to achieve barrier-free communication between hearing impairment patients and healthy hearing people,so that the computer can play a more effective role in human-computer interaction,which can reflect the meaning of body language more directly through the computer,and change the quality of life of the hearing impaired.In recent years,with the rapid development of science and technology,computer has become an irreplaceable tool,and has become an indispensable part of life and production.As a branch of the interaction between human and computer,the visual and vivid sign language recognition has become a hot topic in the field of scientific research.Therefore,the gesture recognition technology has been more and more important in the communication of life and human-computer interaction.Because of the gesture recognition method of video image processing depending on choosing illumination and background extremly,this article designed a wearable glove.The wearable glove can aquire gestures by space sensor,and preprocess the data collected,optimize gesture features,according to the optimized feature,the improved recognition algorithm is used to track and recognize the gesture.The design uses the MPU6050 gyro sensor to collect the data of 3D acceleration and attitude angle in the space trajectory which is the identification model,and use FLX03 bending motion sensor collecting finger bending data,the left and right hand recognition to the data syncing to the left master processor via Bluetooth,then recognizing gestures by S3C2440 processor and improving the accuracy of the recognition algorithm.The gesture collected by the above algorithm preprocess,gesture segmentation,feature extraction,gesture modeling and recognition.In order to improve the efficiency and recognition rate of sign language translation equipment,this paper proposes an improved HMM algorithm based on the combination of time rule recognition algorithm with hidden Markov identification,and verifies the selected eight custom gestures and four Arabia numbers and four English letters respectively.The simulation data,and then analyzed the simulation results show that after the design goals of the system can realize the expected,proved that the algorithm improves the gesture recognition accuracy and recognition efficiency,provide a theoretical basis for large-scale gesture recognition later. |
