THE 3 rd INTERNATIONAL SCIENTIFIC CONFERENCES OF STUDENTS AND YOUNG RESEARCHERS dedicated to the 99
th
anniversary of the National Leader of Azerbaijan Heydar Aliyev
186
obyekt ayrıca nəzərdən keçirilməklə həm sahəsi, həm də yerləşmə
koordinatları tapıla bilir.
5.
Daha sonra “find” əmrinin köməyi ilə müəyyən şərtlər ilə filtirləmənin
aparılması.
Ədəbiyyat [1] Q.Rüstəmov, "Avtomatik Tənzimləmə Nəzəriyyəsi – MATLAB/Simulinkdə
modelləşdirmə", Bakı, 2012
[2] R.C. Gonzalez və R.E. Woods, " Digital Image Processing (3rd Edition)", Nyu York, 2016
DESİGN OF INTELLIGENT CAR INTERIOR COMPONENTS CONTROL BASED ON SPEECH RECOGNITION Rustam Mammadli Department of Process Automation Engineering Baku Higher Oil School Baku, Azerbaijan rustam.memmedli.std@bhos.edu.az Supervisor: Ph.D, Associate Professor Ali Parsayan Keywords: Speech Recognition, Raspberry Pi, STM32, SPI, Short-time Fourier
transform, Mel-frequency cepstral coefficients, Convolutional neural network.
Thesis As different control techniques are arising, it is a matter of task to make
it more humane. This article aims to design a control system for car
components that can be controlled by speech commands. By using Google
Speech API various existing languages can be integrated, however, different
paths should be followed to implement a system that recognizes commands
in Azerbaijani. The system consists of two main parts; the processing of voice
input obtained from the microphone through Raspberry Pi and control of the
components by STM32. This system helps drivers to focus more on the road
rather than dealing with opening/closing or turning on/off something inside
the car by physically interacting with it. There are 4 commands:
“Aç”, “Ba ğ la”, “Qo ş ”, “Söndür” .
For speech processing, different methods can be used. Firstly, after
applying
