عنوان المقالة:تقنية فعالة لكشف التزوير باستخدام النمط الثنائي المحلي (Hessian and Center Symmetric) وطريقة التحويل An Efficient technique for Image Forgery Detection Using Local Binary Pattern (Hessian and Center Symmetric) and Transformation Method
مهند فاضل جويد | Mohanad Fadhil Jwaid | 3433
نوع النشر
مؤتمر علمي
المؤلفون بالعربي
مهند فاضل جويد , تربتي باراسكار
المؤلفون بالإنجليزي
Mohanad F. Jwaid, Trupti N. Barascar
الملخص العربي
In today’s world image tampering or digital image forgery are very common with different tools and software available, which many time leads to digital crime and has an adverse effect on business and technology sector. A digital image provides a compact and authentic way to convey various secure information. To provide more authenticity to it, here a new passive image forgery detection technique is served and implemented. This proposed method is based on Hessian and Center Symmetric Local Binary Pattern, Principle Component Analysis and Discrete Wave Transform. This feature extraction method is applied on the forge input image which is applicable for identifying forge image using copy-move and splicing approach. The proposed method consists of following steps: firstly preprocessing, and second is feature extraction using Local interest point of each object based on Hessian method and extraction Center-Symmetric local binary pattern features. Thirdly Discrete Wavelet Transform has applied for simulations analysis of low-frequency sub-band and removing noise from the image. Fourthly Principle Component Analysis is applied to feature vectors to reduce its dimension while the efficiency well results. Last step using machine learning technique which requires training and testing phases, for a non-linear problem, as kernel function is used that projects the data to a higher dimensional space where it becomes linearly separable. Expected result of the proposed method will be standard metrics that evaluated the localization performance, the True Positive Rate and True Negative Rate are calculated.
تاريخ النشر
20/03/2018
الناشر
iPGCon 2017
رقم المجلد
رقم العدد
رابط الملف
تحميل (63 مرات التحميل)
الكلمات المفتاحية
Image forgery, Local Binary pattern (LBP), Discrete Wavelet Transform (DWT), Principle Component Analysis (PCA), TPR, TNR
رجوع