Image Copy Detection with Rotating Tolerance

Wu, Ming-Ni; Lin, Chia-Chen; Chang, Chin‐Chen

doi:10.1007/11596448_68

Cited by 8 publications

(6 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…WU: a dynamic center detection mechanism for an elliptical track is employed on the gray image. The method divides the gray image into some elliptical tracks, and then computes the mean and variance value of each elliptical track [2].…”

Section: Resultsmentioning

confidence: 99%

“…However, Kim's method fails to deal with copies of 90 or 270 rotation and with only a minor rotation. To solve this weakness, an elliptical track division strategy was proposed, and the mean and variation value of each elliptical track were combined to become the signature set of the image in [2]. The DCT coefficients sign for CBIR was firstly proposed in [14], which is fast and simple.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Image copy detection using a robust gabor texture descriptor

Liu

Jiang

et al. 2009

Proceedings of the First ACM Workshop on Large-Scale Multimedia Retrieval and Mining

View full text Add to dashboard Cite

In this paper, we propose a novel scale and rotation invariant Gabor texture descriptor for content-based image copy detection. Firstly, the Gabor filters with different orientations and scales using a fixed-size window are constructed. Secondly, the Discrete Fourier Transform is implemented on each image. The lowfrequency region of an image with the fixed size is extracted, and is filtered by the Gabor filters with the same fixed-size window. Thirdly, a statistical feature of these Gabor-filtered images is extracted as a scale invariant Gabor texture descriptor. At last, the circular shift method is employed on the scale invariant feature to achieve the rotation invariance. Experimental results demonstrate that the proposed feature is effective to resist various types of image transformations, such as scaling,

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Image copy detection using a robust gabor texture descriptor

Liu

Jiang

et al. 2009

Proceedings of the First ACM Workshop on Large-Scale Multimedia Retrieval and Mining

View full text Add to dashboard Cite

show abstract

“…To detect the images that are manipulated with different rotated angles, an improved content-based image copy detection scheme is proposed by Wu et al [12]. They found that after the image was rotated, some of the edge features would move to other locations.…”

Section: Content-based Image Copy Detectionmentioning

confidence: 99%

“…However, it achieved a poor performance in the processing of rotated digital images, which can only detect the manipulation with a 180 • rotation. To detect the digital images with different rotation angles, Wu et al [12] proposed a scheme by extracting and comparing image feature values in the Y channel of the YUV (luminance, blue and red chrominance) color model, and could successfully detect the infringed images with rotation of 90 • or 270 • . However, it is still not able to effectively detect infringed images with others rotation angles.…”

Section: Introductionmentioning

confidence: 99%

Content-Based Image Copy Detection Using Convolutional Neural Network

et al. 2020

View full text Add to dashboard Cite

With the rapid development of network technology, concerns pertaining to the enhancement of security and protection against violations of digital images have become critical over the past decade. In this paper, an image copy detection scheme based on the Inception convolutional neural network (CNN) model in deep learning is proposed. The image dataset is transferred by a number of image processing manipulations and the feature values in images are automatically extracted for learning and detecting the suspected unauthorized digital images. The experimental results show that the proposed scheme takes on an extraordinary role in the process of detecting duplicated images with rotation, scaling, and other content manipulations. Moreover, the mechanism of detecting duplicate images via a convolutional neural network model with different combinations of original images and manipulated images can improve the accuracy and efficiency of image copy detection compared with existing schemes.

show abstract

“…Hasil penelitian diperoleh dari pengujian terhadap 5 kelompok citra yang telah dipalsukan dengan teknik copy-move dan juga dimodifikasi dengan operasi rotasi [8] dan histogram ekualisasi. Pengujian juga dilakukan dengan menggunakan blok pendeteksian berukuran 64 × 64, 32 × 32 dan 16 × 16.…”

Section: Hasil Dan Pembahasanunclassified

Deteksi Pemalsuan Citra dengan Teknik Copy-Move Menggunakan Metode Ordinal Measure dari Koefisien Discrete Cosine Transform

Zulfan

Arnia

Muharar

2016

JNTE

View full text Add to dashboard Cite

Abstrak-Artikel ini membahas tentang metode baru untuk deteksi citra palsu yang dihasilkan dari teknik copy-move. Teknik copy-move merupakan salah satu teknik pemalsuan citra dengan cara mengambil objek tertentu dari citra asli dan menambahkannya pada citra tersebut dengan tujuan untuk menambah jumlah atau merubah objek yang sama pada citra asli. Penelitian ini bertujuan untuk mendeteksi citra palsu yang dihasilkan oleh teknik copy-move dan citra palsu copy-move yang telah dimodifikasi dengan operasi rotasi dan ekualisasi histogram. Fitur deteksi yang digunakan adalah Ordinal Measure dari koefisien Discrete Cosine Transform (OM-DCT). Pendeteksian dimulai dengan membagi citra ke dalam blok berukuran BxB (B = 16x16, 32x32 dan 64x64) dan DCT 2 dimensi dilakukan pada setiap blok tersebut. Jarak fitur citra asli dengan palsu dihitung dengan persamaan jarak Ecluidian dan setiap fitur yang memiliki jarak lebih kecil atau sama dengan nilai threshold (T) menurut pengamatan akan ditandai sebagai bagian yang dipalsukan. Hasil pendeteksian menunjukkan bahwa ada blok-blok yang terdeteksi pada objek citra yang di-copy-move baik pada citra palsu copy-move yang tidak dimodifikasi ataupun yang telah dimodifikasi dengan operasi rotasi dan ekualisasi histogram. Jumlah blok yang ditemukan pada objek copy-move bervariasi sesuai ukuran blok pendeteksian yang digunakan. Kata kunci : Discrete Cosine Transform (DCT), ordinal measure dari koefisien DCT, copy-move, rotasi, ekualisasi histogram.Abstract-This article discusses a new method for the detection of forgery images generated by copy-move technique. Copy-move technique is one of image forgery techniques which taking a particular object from its original image and add it on that image for the purpose of increasing the number of or changing the same object in the original image. This study aims to detect the forged image generated by the copy-move techniques and copy-move forged image that has been modified by the rotation operation and histogram equalization. Detection feature used is Ordinal Measure of Discrete Cosine Transform coefficient (OM-DCT). Detection starts with division of the image into a block size of BXB (B = 16x16, 32x32 and 64x64) and two-dimensional DCT was performed to each of blocks. The feature distance from the original to the fake image, was calculated by the Euclidian distance and each feature has a distance of less than or equal to the threshold value (T) according to the observations will be marked as a forged part. The results show that there are blocks detected on the copy-move image, whether on the unmodified copy-move forge image or those which modified by the rotation operation and histogram equalization. The number of blocks that are found in the copy-move object varies according to the size of the detection block used.

show abstract

Image Copy Detection with Rotating Tolerance

Cited by 8 publications

References 6 publications

Image copy detection using a robust gabor texture descriptor

Image copy detection using a robust gabor texture descriptor

Content-Based Image Copy Detection Using Convolutional Neural Network

Deteksi Pemalsuan Citra dengan Teknik Copy-Move Menggunakan Metode Ordinal Measure dari Koefisien Discrete Cosine Transform

Contact Info

Product

Resources

About