Rhys B. Sanchez, Jose Angelo C. Esteves, N. Linsangan
{"title":"Effects of ESRGAN in Sugar Apple Ripeness Detection","authors":"Rhys B. Sanchez, Jose Angelo C. Esteves, N. Linsangan","doi":"10.1109/ICCAE56788.2023.10111452","DOIUrl":null,"url":null,"abstract":"In using the CNN algorithm, detecting and classifying objects will output classification errors in images throughout its usage. Presently, using SRGAN was not tested whether it has an effect on the use of CNN in determining sugar apple ripeness. The researchers used Enhanced SRGAN to improve the quality of sugar apple images taken. Images taken from sugar apples were compared to images stored in the dataset of the device using CNN, which then tells the ripeness stage of the sugar apple image. The same set of images captured is enhanced using ESRGAN to compare if there will be an effect on the results of the system using CNN. The researchers saw that better resolution and quality of the images could produce better results based on the data collected. Images without ESRGAN saw an accuracy of 84.00% and with a confidence of 49.57%. Enhanced images using ESRGAN produced more promising results with an accuracy of 92.00% and a confidence of 52.61% compared to normal images.","PeriodicalId":406112,"journal":{"name":"2023 15th International Conference on Computer and Automation Engineering (ICCAE)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 15th International Conference on Computer and Automation Engineering (ICCAE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCAE56788.2023.10111452","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In using the CNN algorithm, detecting and classifying objects will output classification errors in images throughout its usage. Presently, using SRGAN was not tested whether it has an effect on the use of CNN in determining sugar apple ripeness. The researchers used Enhanced SRGAN to improve the quality of sugar apple images taken. Images taken from sugar apples were compared to images stored in the dataset of the device using CNN, which then tells the ripeness stage of the sugar apple image. The same set of images captured is enhanced using ESRGAN to compare if there will be an effect on the results of the system using CNN. The researchers saw that better resolution and quality of the images could produce better results based on the data collected. Images without ESRGAN saw an accuracy of 84.00% and with a confidence of 49.57%. Enhanced images using ESRGAN produced more promising results with an accuracy of 92.00% and a confidence of 52.61% compared to normal images.