Praseptiyana, Winda Ika (2020) Pemanfaatan Ciri Gray Level Co-occurrence Matrix (GLCM) Untuk Deteksi Melasma Pada Citra Wajah. Sarjana thesis, Universitas Brawijaya.
Abstract
Kulit adalah organ tubuh terluar manusia dengan sensitifitas tertinggi dari lingkungan luar yang dapat menimbulkan penyakit kulit. Penyakit kulit pada negara beriklim tropis seperti Indonesia adalah melasma. Melasma diakibatkan penggunaan kosmetik berlebihan yang bersentuhan dengan sinar ultraviolet. Jika dibiarkan dapat merusak sel kulit, merusak DNA dan berisiko timbulnya kanker kulit. Pemeriksaan on site dilakukan dokter kulit mengandalkan pemeriksaan visual dan anamnesis, yang tidak menutup kemungkinan menimbulkan analisis dan diagnosis yang kurang akurat. Sehingga penderita memilih melakukan selfcare. Namun, selfcare dapat menyebabkan melasma semakin parah apabila salah mengidentifikasi. Oleh karena itu, dibutuhkan sistem deteksi untuk membantu mengidentifikasi melasma secara otomatis. Menggunakan data sejumlah 20 citra wajah terbagi 16 citra latih dan 4 citra uji. Citra wajah diolah melalui pemotongan citra secara non-overlapping sliding window untuk mendapatkan citra window, lalu dikonversi ke citra grayscale. Menggunakan metode Gray Level Co-occurrence Matrix (GLCM) sebagai ekstraksi tekstur dengan arah sudut kombinasi 0°, 45°, 90°, 135° dan nilai jarak ketetanggaan d=1,2,3. Kemudian mencari nilai fitur GLCM yakni kontras, homogenitas, energi dan entropi. Berdasarkan nilai dari fitur GLCM yang didapatkan selanjutnya dilakukan tahap klasifikasi menggunakan metode K-Nearest Neighbour (KNN) dengan nilai K=5. Penelitian ini berhasil pada pengujian citra window, persentase terbaik 98% dengan ukuran window 200x200 piksel melalui arah sudut kombinasi 0°+45°+90° dan jarak ketetanggaan d=2.
English Abstract
Skin is the outermost human organ with the highest sensitivity from external environment, it can cause skin diseases. Skin disease in tropical countries like Indonesia is melasma. Melasma is caused by excessive use of cosmetics and contact with ultraviolet light. If allowed to damage skin cells, damage DNA and risk of skin cancer. Examination on site by a dermatologist relies on visual examination and history taking, which does not rule out the possibility of inaccurate analysis and diagnosis. Therefore, patients choose to do selfcare. However, selfcare can cause melasma to get worse if it is misidentified. With that result, a detection system is needed to help identify melasma automatically. Using data of 20 face images divided into 16 training images and 4 testing images. Face images come in processed cropping images with non-overlapping sliding window to get window images, then converted to grayscale images. Using the Gray Level Co-occurrence Matrix (GLCM) method as texture extraction with combination angle is 0°, 45°, 90°, 135° and neighboring distance is d=1,2,3. Use of GLCM features are contrast, homogeneity, energy and entropy for find value of GLCM features. Value of the GLCM features obtained from the classification using the K-Nearest Neighbor (KNN) method with value is K=5. This research suceessin testing of window images, the best percentage was 98% with window size of 200x200 pixels, combination angle is 0°+45°+90° and adjacency distance is d=2.
Other obstract
-
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FILKOM/2020/206/052003397 |
| Uncontrolled Keywords: | penyakit kulit, melasma, ekstraksi ciri, GLCM, KNN, skin disease, melasma, feature extraction, GLCM, K |
| Subjects: | 000 Computer science, information and general works > 006 Special computer methods > 006.4 Computer pattern recognition |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 02 Oct 2020 02:52 |
| Last Modified: | 12 Apr 2023 07:20 |
| URI: | http://repository.ub.ac.id/id/eprint/180946 |
![[thumbnail of DALAM MASA EMBARGO]](http://repository.ub.ac.id/style/images/fileicons/text.png) |
Text (DALAM MASA EMBARGO)
Winda Ika Praseptiyana (2).pdf Restricted to Registered users only until 31 December 2023. Download (18MB) |
|
Text (DALAM MASA EMBARGO)
Winda Ika Praseptiyana (2).pdf Restricted to Registered users only until 31 December 2023. Download (18MB) |
Actions (login required)
 |
View Item |