DIGITAL WATERMARKING PADA CITRA TERKOMPRESI ANIMATED GIF DENGAN METODE BIT-PLANE COMPLEXITY SEGMENTATION (BPCS)

• Main Author: Hidayanti, Fitri (Author)
• Format: Book
• Published: 2016-06-17.
• Subjects: Thesis; NonPeerReviewed
• Online Access: Link Metadata
• Notes: http://repository.upi.edu/28068/1/S_PLB_1007983_Title.pdf; http://repository.upi.edu/28068/2/S_PLB_1007983_Abstract.pdf; http://repository.upi.edu/28068/3/S_PLB_1007983_Table_of_content.pdf; http://repository.upi.edu/28068/4/S_PLB_1007983_Chapter1.pdf; http://repository.upi.edu/28068/5/S_PLB_1007983_Chapter2.pdf; http://repository.upi.edu/28068/6/S_PLB_1007983_Chapter3.pdf; http://repository.upi.edu/28068/7/S_PLB_1007983_Chapter4.pdf; http://repository.upi.edu/28068/8/S_PLB_1007983_Chapter5.pdf; http://repository.upi.edu/28068/9/S_PLB_1007983_Bibliography.pdf; http://repository.upi.edu/28068/10/S_PLB_1007983_Appendix.pdf
• Keywords: QA75 Electronic computers. Computer science; T Technology (General)

Salah satu solusi untuk menangani kasus-kasus terkait hak cipta adalah dengan menerapkan teknologi Digital Watermarking. Format GIF merupakan salah satu format citra digital yang banyak dijumpai dan umum dipakai di internet karena ukurannya yang cukup kecil dan dapat menampilkan animasi sederhana. Metode kompresi yang digunakan file GIF adalah varian dari teknik kompresi LZW (Lempel-Ziv-Welch). Ada banyak metode yang digunakan untuk digital watermarking, penelitian ini mengimplementasikan metode Bit-Plane Complexity Segmentation (BPCS), salah satu teknik steganografi yang memanfaatkan perhitungan kompleksitas pada tiap bit-plane dalam menyisipkan data rahasia. Penelitian ditujukan untuk menganalisis kecocokan metode steganografi BPCS dalam menyisipkan watermark di dalam sebuah citra digital terkompresi berformat GIF animasi. Berdasarkan hasil pengujian nilai PSNR untuk mengukur kualitas watermarked-image, dari total 30 data uji semua citra watermarked berhasil menghasilkan nilai PSNR diatas 30dB. Rata-rata nilai PSNR yaitu sebesar 36,79dB sehingga metode BPCS dapat dikatakan bekerja dengan baik pada file animated GIF indexed color (mutu citra penampung tidak jauh berubah) dan memenuhi faktor fidelity. Berdasarkan pengujian faktor robustness, metode BPCS pada animated GIF tidak robust terhadap manipulasi penajaman kontras, rotasi, vertical flip, grayscale dan biner. Dari penambahan 5 operasi manipulasi gambar tersebut, watermark yang telah disisipi tidak berhasil terungkap kembali. Rata-rata tingkat recovery terhadap manipulasi gambar yaitu 0%, sehingga metode ini rentan terjadi kerusakan gambar terhadap operasi manipulasi. Selain itu dilakukan juga analisis pengujian terhadap lama waktu proses penyisipan dan pengungkapan watermark. Berdasarkan hasil pengujian disimpulkan bahwa kecepatan proses embedding dan extracting dapat dipengaruhi oleh panjang pesan watermark, jumlah frame dan dimensi gambar (panjang, lebar). ;---One of the solution to protect copyright is by implementing Digital Watermarking technique. GIF is one of the multimedia format which is commonly used on the internet because of its small size and can display simple animations. Compression method used by GIF file is a variant of the LZW compression technique (Lempel-Ziv-Welch). There are some methods used for digital watermarking on the image document. This research implements the Bit-Plane Complexity Segmentation (BPCS) method, one of the steganography techniques that utilize calculation of complexity at each bit-plane in concealing a message or information. This research intends to analyze the suitability of the BPCS steganography method to insert a watermark in a digital compressed image animated GIF. Based on the test results to measure the quality of watermarked image, from 30 image tests can be obtained that all images have PSNR value above 30dB. The average value of PSNR is 36,79dB which can be concluded that BPCS is work well with animated GIF (image quality is not much changed) and occupy the fidelity factor. Based on the test of robustness factor, BPCS method in animated GIF is not robust against manipulation image like contrast enhancement, rotation, vertical flip, grayscale and binary. From 5 operation of manipulating image, none of the watermark can be recovered. The average rate of recovery against the manipulation image is 0%, so this method is susceptible to damage image against image manipulation. The research also analyzes the embedding and extracting time proccess. Based on the test, can be concluded that the speed of embedding and extracting time is influenced by the length of the watermark message, the number of frames and image dimensions (length, width).