Please use this identifier to cite or link to this item:
https://dspace.univ-ouargla.dz/jspui/handle/123456789/35014
Title: | Soil recognition and features measurement using AI |
Authors: | Amran, Laila Ferhi, Aya |
Keywords: | Artificial Intelligence Deep Learning ,CNN Proposed CNN VGG16 ResNet Mobile Net |
Issue Date: | 2023 |
Publisher: | UNIVERSITY OF KASDI MERBAH OUARGLA |
Abstract: | Soilplaysanimportantroleinthequalityofagriculturalcrops,especiallysinceweareinanera inwhichtheagriculturalsectoroccupiesagreatimportancenotonlyfromthesideoftheeconomy,butalsofromthesideofachievingfoodsufficiency,soifthetypeofthesoilisnotappropriate,thentheproductwillnotbeoftherequiredquality. Inourresearch,wewantedtohelpevery farmerandpeasantandeveryoneinterestedinthisfieldbydevelopingawebsitethatallowssoil classification(initiallyfivetypes)byincludingapictureofthesoiltoknowitstype. TheclassificationprocesswascarriedoutusingdeeplearningexactlyProposedCNNmodel,whichhadthe highestaccuracy(86%)aftercomparingitwith3othermodels,VGG(80%),ResNet(79%),and MobileNet(76%). Intheend,aftertheclassificationisdone,themostimportantcharacteristics (sixcharacteristics)ofthatsoilarepresented,whitchistemperature,Ph,Porosity,texture,color ,density. |
URI: | https://dspace.univ-ouargla.dz/jspui/handle/123456789/35014 |
Appears in Collections: | Département d'informatique et technologie de l'information - Master |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.