Centre of Excellence in Cognitive Science - CECoS

College of Research Methodology and Cognitive Science (RMCS), Burapha University

 

  Home   Event    People    Research   Project  ○  Publication  ○  Collaboration  ○  Contact Us

 

Senior engineering projects done in 2010

Senior engineering projects done in 2009

Senior engineering projects done in 2008

Copyright © All rights reserved.

 

" Nondestructive Maturity Detector of Fruits by using Knock Sound Processing "

Project Team: Peerapong Somboonyod, Theerapon Paritwanon, and Warunee Bundit

URL: http://www.rmcs.buu.ac.th/cecos/project/durian  (Thai)

 

Problem: Thailand is the land of agricultural products in which the quality of fruit is importance for product grading. In 2006, Thailand can export agricultural products about 155,000 Tons or about 3.2 Billion Baht. Since the knocking method is a traditional knowledge of Thai people for long time, this method can be able to check whether the fruit is maturity or immaturity. However, this method requires more experiences and only experts can do.

Basic Idea: This project proposes a simple nondestructive method for determining maturity of fruits by using knock-sound processing. The basic idea of the proposed method is on the basis of a force and a position of knocking can be set. Based on this assumption, we can design automatic machine for knocking and recording the knock sound. The knocked sound will be analyzed based on the signal processing techniques.

 

Working process of the proposed method

 

Prototype of automatic knocking machine

Method: The designed knocking machine consists of a bidirectional DC motor 8.33 A, H-bridge circuit, microcontroller PIC18F4550, and movable knock-arm connected with the DC motor by a machine belt. Microcontroller is used as a central control unit to control angular velocity and direction of knock-arm. During experiment, a sound of knocking was recorded through the condenser microphone which to be considered as an impulse response of fruit. To process such sound of knocking, the recorded sound was converted to the digitized signal at sampling rate of 44,100 Hz with 16-bit resolution and a total length of whole sampling 4,096 points. The sampled data will be processed as a small segment or called windowing technique. To determine the rate of change, we proposed 128-point windowing technique with computation of standard deviation (SD) value. This process is repeated over total length of data therefore we then obtain 32 points per a time of knocking. We can specific threshold value between maturity and immaturity of fruit by applying probability density function of the distribution of SD values.

 



FDM Software Version 1.0 used as a knock sound processor

 

Experimental Results: In the experiment, 10 pineapples and 10 durians (of “Monthong” variety) to be tested for showing the effectiveness of the proposed method. The proposed method can be able to classify the maturity of fruits by the average rates of 95.5% for pineapples, 85.0% for  the weight of durian 3.5-4.0 kg. and 88.7% for the weight of durian 4.0-4.5 kg.

Publication: T. Paritwanon, P. Somboonyod, W. Bundit, and M. Phothisonothai, "Non-Destructive Fruit Maturity Determination by using Knock Sound Processing" The 10th Annual Conference of Thai Society of Agricultural Engineering on Innovations in Agricultural, Food and Renewable Energy Productions for Mankind (TSAE2009), pp.344-349, April 2009.

 

 

" Development of Computer Software for Pathological Analyzing of Thai Language Learners "

Project Team: Nalinee Chanamool and Benyada Unhalekajit

URL: http://www.rmcs.buu.ac.th/cecos/project/speech/A.html (Thai)

 

Problem: Since Thai tone is unique comparing with other languages because the different part of consonant phonemes from the vowel. It plays key role in pronouncing Thai language especially in the specific tones, segmentation of each consonant phoneme from the vowel is required for analyzing Thai speech signal. The problem of Thai pronunciation has been recognized as the nation issue and there was the Thai Language Clinic for Youth.
 
 

System overview of the phonetic analyzing method

 

 Typical example of specific tone segmentation using the proposed method

 

Basic Idea: This project proposes the development of computer software for phonetic analyzing based on the speech processing techniques. We named this project as the "PATHAI (Phonetic Analyzer for THAI language learners)" software. The PATHAI software supports not only Thai speakers but also foreigners as well, they can be able to practice themselves the pronunciation of specific Thai words.

 


Graphs plot the values of two parameters

 

Method: There are three main processes of the proposed method; 1) preprocessing, 2) feature extraction and 3) classification. This project selected two parameters that are zero-crossing rate (ZCR) and time duration (TD) of phonemes for creating the template database. The speech signal of correct pronouncing is recorded, and then ZCR and TD of each word will be categorized as the templates.

 


PATHAI Software Version 1.0 (developed by using Delphi language)

 

 

Analyzing the recorded speech by PATHAI

 

Experimental Results: The 5 females (aged 20 - 25 years) and 5 males (aged 20 - 25 years)  volunteered to participate in this experiment. The results show that the method can archive the accuracy rates of 61.1% - 90.0% approximately.

Publication:  N. Chanamool, B. Unhalekajit, and M. Phothisonothai, "Development of Computer Software for Phonetic Analyzing of Thai Language Speakers" The National Computer Science and Engineering Conference (NCSEC2009), pp.445-450, Bangkok, Thailand, 2009.

 

Copyright © 2009-11 CECoS, College of Research Methodology and Cognitive Science, Burapha University, Chonburi 20131 Thailand.