No abstract

No abstract

Seiring perkembangan teknologi dilakukan otomatisasi deteksi kanker kulit melalui citra dermoscopy. Pengambilan informasi fitur citra dermoscopy terganggu dengan outlier dan overfitting, karena faktor jenis kulit, penyebaran kanker yang tidak merata atau kesalahan sampling. Penelitian ini mengusulkan deteksi kanker kulit melanoma dengan mengintegrasikan metode fuzzy K-Nearest Neighbour (FuzzykNN), Lp-norm dan Linear Discriminant Analysis (LDA) untuk mengurangi outlier dan overfitting. Masukan berupa citra warna RGB yang dinormalisasi menjadi RGBr. Reduksi dimensi dengan LDA menghasilkan fitur dengan nilai eigen paling menonjol. LDA pada penelitian ini menghasilkan dua fitur paling menonjol dari 141 jenis fitur, yaitu wilayah tumor dan minimum wilayah tumor channel R. Kemudian dilakukan klasifikasi FuzzykNN dan metode pengukur jarak Lp-norm. Penggunaan metode LDA dan Lp-norm dalam proses klasifikasi ini mengatasi terjadinya overfitting. Akurasi yang dihasilkan metode LDA-fuzzykNN Lp Norm, yaitu 72% saat masing-masing nilai p dan k = 25. Metode gabungan ini terbukti cukup baik dari pada metode yang dijalankan terpisah.

Fitur yang digunakan untuk mengenali jenis daun meliputi bentuk, warna, dan tekstur. Tidak semua jenis fitur perlu digunakan untuk melakukan komputasi hasil ektraksi, namun perlu diseleksi beberapa fitur yang paling berpengarauh dalam sistem temu kembali citra daun. Teknik seleksi fitur Correlation based Featured Selection (CFS) digunakan untuk melakukan pemilihan fitur berdasarkan korelasi antar fitur, sehingga dapat meningkatkan performa dari sistem temu kembali citra daun. Jenis seleksi fitur yang digunakan diantaranya menggunaka CFS, CFS dengan Genetic Search (GS), dan chi square. Analisis keterkaitan korelasi antar fitur melalui seleksi fitur juga dikombinasikan dengan penggunaan kedekatan dalam menghitung similaritas pada sistem temu kembali. Penggunaan kedekatan dengan Lp norm, manhattan, euclidean, cosine, dan mahalanobis. Hasil penelitian ini menunjukkan nilai temu kembali paling tinggi ketika menggunakan seleksi fitur CFS dengan pengukuran kedekatan mahalanobis.

No abstract

Batik is one of the Indonesian cultures that UNESCO has recognized. Batik has a variety of unique and distinctive patterns that reflect the area of origin of the batik motif. Batik motifs usually have a 'core motif' printed repeatedly on the fabric. The entry of digitization makes batik motif designs more diverse and unique. However, with so many batik motifs spread on the internet, it is difficult for ordinary people to recognize the types of batik motifs. This makes an automatic classification of batik motifs must continue to be developed. Automation of batik motif classification can be assisted with artificial intelligence. Machine learning and deep learning have produced much good performance in image recognition. In this study, we use deep learning based on a Convolutional Neural Network (CNN) to automate the classification of batik motifs. There are two datasets used in this study. The old dataset comes from a public repository with 598 data with five types of motifs. Meanwhile, the new dataset updates the old dataset by replacing the anomalous data in the old dataset with 621 data with five types of motifs. The lereng motif is changed to pisanbali due to the difficulty of obtaining the lereng motif. Each dataset was divided into three ways: original, balance patch, and patch. We used ResNet-18 architecture, which used a pre-trained model to shorten the training time. The best test results were obtained in the new dataset with the patch way of 88.88 % ±0.88, and in the old dataset, the best accuracy was found in the patch way on the test data of 66.14 % ±3.7. The data augmentation in this study did not significantly affect the accuracy because the most significant increase in accuracy is only up to 1.22%.

Object Detection using Haar Cascade Clasifier widely applied in several devices and applications as a medium of interaction between human and computer such as a tool control that utilizes the detection of eye movements. Obviously speed and precision in the detection process such as eyes, has an effect if implemented on a device. If the eye could not detect accurately, controlling device systems could reach bad detection as well. The proposed method can be used as an approach to detect the eye region of eye based on haar classifier method by means of modifying the direction of sliding window. In which, it was initially placed in the middle position of image on facial area by assuming the location of eyes area in the central region of the image. While the window region of conventional haar cascade scan the whole of image start from the left top corner. From the experiment by using our proposed method, it can speed up the the computation time and improve accuracy significantly reach to 92,4%.

No abstract