Salma Mohamed, Shadwa Khaled, Maha Ehab, Bassant Mohamed
4 May 2022
Electronic tongues are a new, promising technology that has already proven to be a useful tool for assessing and predicting the taste of beverages and associated items. The main idea of this project is to classify any milk sample into pure or adulterated. Our method is much easier than human taste as human taste can’t always classify the milk if it’s pure or adulterated, and also to study the tastes of the tongue and classify the liquids (milk) to detect if it’s edible or not by containing adulterants and it will help in increasing the production as it saves money and effort and the machines are always faster than a human which will lead to faster production. Classification will be done by using a PH meter, conductivity, the electrodes which are used in the PH meter, and conductivity. Milk will be measured in different concentrations. Some machine learning algorithms would be used to test the accuracy such as Support Vector Machine (SVM). Also, deep learning algorithms would be used as Multilayer Perceptron’s (MLPs).
The main purpose of this SDD document is to illustrate and facilitate analysis, planning, and implementation for our graduation project (Electronic Tongue). This document will explain in detail the system software (mobile application) and how our results will appear on the application. The software implementation includes different machine learning algorithms and methods used. The hardware used is the PH meter, conductivity, and the electrodes that are in the PH meter and conductivity. E-Tongue is mainly classifying liquids according to their different tastes. But, Our goal is to classify the milk as to whether it’s adulterated or pure. The python code will be connected by the mobile application by using an API to make it easy to take the user’s input and test it with the used model to know if the milk is adulterated or pure. Our target audience is the Food Control Authority, Ministry of Health, commercial fraud authorities, and the faculties of pharmacy (according to the used hardware).
This system will cause a great revolution in the production field. The electronic tongue has several advantages, including reduced detection fatigue, and its results are comparable and trustworthy. Our proposed project will help in saving money as the system will replace the connoisseurs and needed materials for milk adulteration tests. So, it won’t be essential for the places to hire many people. Moreover, it will save more time as the machine is faster than the human and more accurate. The ET system will be used to classify the milk as if it’s adulterated or pure.
The Software design document will describe in detail the architecture of the Electronic Tongue System. In the first part of the SDD, there is an introduction that will discuss the purpose of the document, the scope, and the overview. The second part talks about the overview of the system, the scope of the system, the objectives, and the timeline plan. In the third part, we discussed the system’s design viewpoints, pattern use viewpoint, algorithms viewpoint, interaction viewpoint, and interface point. Moreover, the fourth part talks about the dataset which is used in the project in detail. The fifth part talks about the human interface design which is the software part. The sixth part talks about the requirement matrix (which includes the test plans). The last part of the document shows our GitHub repository and the appendices used.
1.4 Intended audience
Our intended audience are: Food Control Authority, Ministry of Health, Commercial fraud authorities, The faculties of pharmacy (according to the used hardware).