A. Husein, Abyan Ramzi, N. Muzakki, Rahmawati Hasanah
{"title":"Quameaty:利用InceptionV3模型,基于图像处理的天然鸡肉质量检测应用","authors":"A. Husein, Abyan Ramzi, N. Muzakki, Rahmawati Hasanah","doi":"10.31544/jtera.v7.i1.2022.107-114","DOIUrl":null,"url":null,"abstract":"model (COCO) model top 5 93,3% dibandingkan 90% , 50% Abstract Nowadays, chicken meat has become a good source of animal protein for consumption and easy to obtain. However, in the process of obtaining it, fraudulent practices are often found, such as tiren chicken meat that is still being sold, or chicken meat that has been mixed with meat that is not worth selling. Therefore, we need a tool or application that is able to detect the quality of raw chicken meat. The purpose of this study is to create a tool that is useful in detecting the quality of raw chicken meat by utilizing image processing using the InceptionV3 model and named Quameaty. This tool was developed using the Python programming language. The InceptionV3 model is an excellent convolutional neural network training model and has been pre-trained on the Common Objects in Context (COCO) dataset of 328,000 images with 81 different classes. This model has a very high level of accuracy as a pre-trained model with a top 5 accuracy value of 93.3% and a relatively fast computation time when compared to its predecessor model. The resulting training model is embedded in an Android application which can be easily and tends to be flexible to be used in detecting the quality of raw chicken meat. The results of the study were divided into two classes, namely fresh and rotten, and showed that the prediction of the quality of raw chicken meat went well with the test metric values that had reached more than 90% at two threshold values, namely 50% and 75%.","PeriodicalId":17680,"journal":{"name":"JTERA (Jurnal Teknologi Rekayasa)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2022-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Quameaty: Aplikasi Pendeteksi Kualitas Daging Ayam Mentah Berbasis Pengolahan Citra Menggunakan Model InceptionV3\",\"authors\":\"A. Husein, Abyan Ramzi, N. Muzakki, Rahmawati Hasanah\",\"doi\":\"10.31544/jtera.v7.i1.2022.107-114\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"model (COCO) model top 5 93,3% dibandingkan 90% , 50% Abstract Nowadays, chicken meat has become a good source of animal protein for consumption and easy to obtain. However, in the process of obtaining it, fraudulent practices are often found, such as tiren chicken meat that is still being sold, or chicken meat that has been mixed with meat that is not worth selling. Therefore, we need a tool or application that is able to detect the quality of raw chicken meat. The purpose of this study is to create a tool that is useful in detecting the quality of raw chicken meat by utilizing image processing using the InceptionV3 model and named Quameaty. This tool was developed using the Python programming language. The InceptionV3 model is an excellent convolutional neural network training model and has been pre-trained on the Common Objects in Context (COCO) dataset of 328,000 images with 81 different classes. This model has a very high level of accuracy as a pre-trained model with a top 5 accuracy value of 93.3% and a relatively fast computation time when compared to its predecessor model. The resulting training model is embedded in an Android application which can be easily and tends to be flexible to be used in detecting the quality of raw chicken meat. The results of the study were divided into two classes, namely fresh and rotten, and showed that the prediction of the quality of raw chicken meat went well with the test metric values that had reached more than 90% at two threshold values, namely 50% and 75%.\",\"PeriodicalId\":17680,\"journal\":{\"name\":\"JTERA (Jurnal Teknologi Rekayasa)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-06-26\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"JTERA (Jurnal Teknologi Rekayasa)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.31544/jtera.v7.i1.2022.107-114\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"JTERA (Jurnal Teknologi Rekayasa)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.31544/jtera.v7.i1.2022.107-114","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Quameaty: Aplikasi Pendeteksi Kualitas Daging Ayam Mentah Berbasis Pengolahan Citra Menggunakan Model InceptionV3
model (COCO) model top 5 93,3% dibandingkan 90% , 50% Abstract Nowadays, chicken meat has become a good source of animal protein for consumption and easy to obtain. However, in the process of obtaining it, fraudulent practices are often found, such as tiren chicken meat that is still being sold, or chicken meat that has been mixed with meat that is not worth selling. Therefore, we need a tool or application that is able to detect the quality of raw chicken meat. The purpose of this study is to create a tool that is useful in detecting the quality of raw chicken meat by utilizing image processing using the InceptionV3 model and named Quameaty. This tool was developed using the Python programming language. The InceptionV3 model is an excellent convolutional neural network training model and has been pre-trained on the Common Objects in Context (COCO) dataset of 328,000 images with 81 different classes. This model has a very high level of accuracy as a pre-trained model with a top 5 accuracy value of 93.3% and a relatively fast computation time when compared to its predecessor model. The resulting training model is embedded in an Android application which can be easily and tends to be flexible to be used in detecting the quality of raw chicken meat. The results of the study were divided into two classes, namely fresh and rotten, and showed that the prediction of the quality of raw chicken meat went well with the test metric values that had reached more than 90% at two threshold values, namely 50% and 75%.