Sistem Rekognisi Menggunakan Metode Local Binary Pattern Dan Support Vector Machine Untuk Mendeteksi Ruang Kosong Tempat Parkir Luar Ruangan

Mochamad Rakha Luthfi Fahsya, Febryanti Sthevanie, Kurniawan Nur Ramadhani

Abstract

Abstrak

Pertambahan volume kendaraan menyebabkan kuota parkir disebuah tempat parkir berkurang dan terjadinya kepadatan pada slot tempat parkir yang menyebabkan kesulitan pengunjung dalam mencari slot tempat parkir kosong, yang akan membuat waktu dalam melakukan pencarian tempat parkir terbuang. Peneliti menggunakan metode Gamma Correction, Gaussian Blur, Local Binary Pattern Rotation Invariant Uniform dan Support Vector Machine dengan setelan parameter Gaussian Blur menggunakan tanpa kernel, Gamma Correction menggunakan nilai gamma = 15, Local Binary Pattern Rotation Invariant Uniform P=8 dan R=1 dan Support Vector Machine kernel RBF mendapatkan nilai akurasi sebesar 99.52% lebih baik dibandingkan jurnal [3] dengan metode ð‘³ï¿½ï¿½ð’– akursi 98.90% dan ð‘³ð‘©ï¿½ï¿½ï¿½ð’Šð’–��dengan akurasi 82.78% .
Kata Kunci: Local Binary Pattern Rotation Invariant Uniform, Support Vector Machine, Gaussian
Blur


Abstract

The increase in vehicle volume causes the parking quota in a parking space to decrease and decreases in the parking space slot which causes difficulty for visitors to find a parking space slot, which will make time in searching for a parking space wasted. Researchers will use the Gamma Correction, Gaussian Blur, Uniform Rotation Local Binary Pattern and Support Vector Machine method using parameters Gaussian Blur using without kernel, Gamma Correction using gamma value = 15, Local Binary Pattern Rotation Invariant uniform parameters P = 8 and R = 1 and RBF kernel Vector Support Engine with accuracy 99.52%% better than journals [3] with the method ð‘³ï¿½ï¿½ð’– accuracy 98.90% and ð‘³ð‘©ï¿½ï¿½ï¿½ð’Šð’–ðŸ with an accuracy of 82.78% .
Keywords: Local Binary Pattern Rotation Invariant Uniform, Support Vector Machine, Gaussian Blur

Full Text:

PDF

Refbacks

  • There are currently no refbacks.
max_upload :0