?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=Pengenalan+Tulisan+Tangan+Alfabet+Menggunakan+Scale-Invariant+Feature+%0D%0ATransform+(SIFT)+%0D%0A&rft.creator=Daniel+Argado+Simanjuntak+%2C+1417051030&rft.subject=005+Pemrograman+komputer%2C+program+dan+data&rft.description=This+research+is+one+of+the+further+research+developments+using+the+SIFT+method+%0D%0Awhich++is++applied++to++Latin++handwriting.++This++study++uses++the++SIFT++method++in+%0D%0Aextracting+features+because+SIFT-based+Descriptors+outperform+other+contemporary+%0D%0Alocal+descriptors+in+the+texture+and+structure+sections%2C+with+a+greater+difference+in+%0D%0Aresults+in+the+texture+section.+This+research+data+consists+of+images+in+each+letter+in+%0D%0Athe+form+of+lower+or+uppercase+letters.+Classification+method+using+SVM+and+10-fold+cross+validation+as+a+comparison+of+accuracy+in+Latin+handwriting+character+%0D%0Arecognition.+The+results+showed+that%3A+(1)+The+extraction+of+the+SIFT+feature+was+%0D%0Asuccessfully+implemented+in+the+Latin+handwriting+character+recognition.+(2)+The+%0D%0Ause+of+the+SIFT+feature+is+good+enough+in+the+introduction+of+Latin+handwriting+%0D%0Acharacters.+(3)+The+highest+accuracy+is+obtained+with+the+use+of+the+LibSVM+Linear+%0D%0AKernel+with+values+reaching+84.13%25+with+a+10-fold+CV+train+and+83.37%25+with+a+%0D%0Anormal++train.+From++this+accuracy%2C+it+can++be+seen+the+Mean++Squared++Error+value%2C+%0D%0Awhich+is+grouped+into+2+types+of+misclassification%2C+(1)+The+shape+of+the+letters+are+%0D%0Asimilar+and+(2)+The+same+number+of+lines+so+that+misclassification+occurs+in+the+%0D%0Adata.+%0D%0AKeywords%3A+D-SIFT%2C+SVM%2C+10-fold+Cross+Validation%2C+Latin+Alfabet%2C+MATLAB.%0D%0A%0D%0APenelitian++ini++merupakan++salah++satu++pengembangan++penelitian++menggunakan+%0D%0Ametode++SIFT++yang++diterapkan++pada++tulisan++tangan++Latin.++Penelitian++ini+%0D%0Amenggunakan+metode+SIFT+dalam+mengekstraksi+fitur+karena+Descriptor+berbasis+%0D%0ASIFT+mengungguli++descriptor+localkontemporer+lainnya+pada+bagian+tekstur+dan+%0D%0Astruktur%2C++dengan++perbedaan++hasil++yang++lebih++besar++pada++bagian++tekstur.++Data+%0D%0Apenelitian+ini+terdiri+dari+citra+pada+masing-masing+huruf+dalam+bentuk+huruf+kecil+%0D%0Aataupun+besar.+Metode+klasifikasi+menggunakan+SVM+dan+10-fold+cross+validation%0D%0Asebagai+perbandingan+akurasi+pada+pengenalan+karakter+tulisan+tangan+Latin.+Hasil+%0D%0Apenelitian++menunjukkan++bahwa%3A++(1)++Ekstraksi++fitur++SIFT++berhasil+%0D%0Adiimplementasikan+pada+pengenalan+karakter+tulisan+tangan+Latin.+(2)+Penggunaan+%0D%0Afitur+SIFT+sudah+cukup+baik+dalam+pengenalan+karakter+tulisan+tangan+Latin.+(3)+%0D%0AAkurasi+tertinggi+didapatkan+pada+penggunaan+Kernel+Linear+LibSVM+dengan+nilai+%0D%0Amencapai+84.13%25+dengan+train+10-fold+CV+dan+83.37%25+dengan+train+normal.+Dari+%0D%0Aakurasi++tersebut++dapat++dilihat++nilai++Mean++Squared++Error%2C++yang++dikelompokkan+%0D%0Akedalam+2+jenis+kesalahan+klasifikasi%2C+(1)+Bentuk+huruf+yang+mirip+dan+(2)+Jumlah+%0D%0Agaris+yang+sama+sehingga+terjadi+misklasifikasi+pada+data+tersebut.+%0D%0AKata+Kunci%3A+D-SIFT%2C+SVM%2C+10-fold+Cross+Validation%2C+Alfabet+Latin%2C+MATLAB.&rft.publisher=FAKULTAS+MIPA&rft.date=2019&rft.type=Skripsi&rft.type=NonPeerReviewed&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F57765%2F1%2FABSTRAK.pdf&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F57765%2F2%2FSKRIPSI%2520FULL.pdf&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F57765%2F3%2FSKRIPSI%2520TANPA%2520BAB%2520PEMBAHASAN.pdf&rft.identifier=++Daniel+Argado+Simanjuntak+%2C+1417051030++(2019)+Pengenalan+Tulisan+Tangan+Alfabet+Menggunakan+Scale-Invariant+Feature+Transform+(SIFT).++FAKULTAS+MIPA%2C+UNIVERSITAS+LAMPUNG.+++++&rft.relation=http%3A%2F%2Fdigilib.unila.ac.id%2F57765%2F