Moblie Computing For Pathological Voice Recognition

Mobile computing, as a result of merits such as universality, rapidity, timeliness and so on, has the important auxiliary function in the treatment and the diagnosis of the Pathology voice on medical clinical. The application in the examination aspect of pathology voice in mobile computing not only may supply patient's condition information at times, moreover to treats provides the powerful basis with the diagnosis.The study method of this article is based on the mobile equipment --cell phone that carries on the voice data with the analysis of frequency domain & time-domain and the multi parameter pattern recognition classifications in the migration computational method and the signal processing method, then it could completes early diagnosis which gets sick to the voice illness. What topic use are the Massachusetts Eyes and Ear Infirmary of American that issues the voice sample database. The main research work is: Realizes the voice audio frequency document record on the cell phone, meanwhile analyze a WAV document of the records document, use autocorrelation sends the extraction tone cycle and definite voice's first parameter"Recognition", then uses Support Vector Machine to carry on the recognition classification, like this it can examine the morbid state voice and the normal voice.Result demonstration: Reaches 89.7% with acoustics parameter to the healthy group and the morbid group's two classified correct recognition rate. For the Pathology voice, this result not only may show the importance of detection of Recognition, but also will strength more feasibility that uses Mobile computing to apply in detection of Recognition.