Effective Deep Features for Image Splicing Detection

Ahmed, Ismail Taha; Hammad, Baraa Tareq; Jamil, Norziana

doi:10.1109/icset53708.2021.9612569

Cited by 15 publications

(8 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These methods rely on hand-made features, and they have many weaknesses that include a high degree of computational complexity, low detection accuracy, and they work on a certain form of forgery [16]. Studies in this field are:  In 2013, [17] Zaho and Guo proposed a method for detecting transcription-type forgery based on discrete cosine transform (DCT) and single value analysis (SVD).…”

Section: Studies That Depend On Traditional Methodsmentioning

confidence: 99%

“…recent studies have addressed this type due to its ability to learn automatically, its accuracy compared to traditional methods for identifying modified regions, and time and effort saving nature required to determine the needed characteristics [16]. These methods have been met with unparalleled success in various fields, such as image processing in forensic digital images and fraudulent images.…”

Section: Studies That Depend On Deep Learningmentioning

confidence: 99%

“…Then, SVM was used, which is from Algorithms that perform linear and non-linear classification based on the kernel is a supervised learning model with data analysis algorithm for future classification.  In 2020, Ahmed and others [16] presented a method for detecting forgery of the digital image splicing type. The method includes converting the image to gray scale and then using Pretrained Alex Net Model, which is a type of pre-trained CNN that is used to extract the important features of the image and these features enter the canonical classifier Correlation Analysis (CCA), which in turn categorizes whether the digital image is fake or real.…”

Section: Studies That Depend On Deep Learningmentioning

confidence: 99%

“…In this paragraph, a summary will be given of the strengths and weaknesses of the algorithms that are being studied, and according to the type of forgery in the images applied to the algorithm within the database used see Tables (1) and (2). [16] Through the study that I presented, I advise researchers to use deep learning models in the event of detecting forgery in digital images, because of their effective role in increasing the accuracy of detection, due to the features that characterize them, including the feature extraction and classification stage, which have been integrated into deep learning models for its ability to learn Features automatically and its ability to distinguish patterns. Even in the case of a small database, we can use pre-trained network models and it is possible to achieve distinct results as in the research [16], which achieved an accuracy of 98% using the Alex-Net model, where it is possible to build a neural network from scratch or it is possible to use trained models In advance, it is superior to traditional methods that depend on extracting features manually, which must pass through three stages: the pre-processing stage, the feature extraction stage and the classification stage.…”

Section: Strengths and Weaknesses Of Previous Algorithmsmentioning

confidence: 99%

See 3 more Smart Citations

A survey on tamper detection techniques for digital images

Taha¹,

Ebraheem²

2022

AL-Rafidain Journal of Computer Sciences and Mathematics

View full text Add to dashboard Cite

Recently, fake and fabricated images that have been manipulated for several purposes, including cosmetic and some for illegal purposes, have spread on social media. And because this is not an easy matter, it has become necessary for researchers in this field to search and investigate the types of images, to verify their authenticity and how to manipulate them. Therefore, the aim of this research is to serve as an assistant to the researcher who wants to enter this field. In this research, a survey of the types of forgery was presented with examples. The most important common methods for detecting forgery were also presented, and the previous studies that contributed to the process of detecting fraud, both traditional detection and deep learning-based detection, were highlighted, while giving the most important strengths and weaknesses of both types. We note from this that detecting the forgery process is a difficult process and takes time to detect the changes that have been made to the image that cannot be detected by the naked eye. In this research, researchers have been urged to go towards deep learning for the purpose of detecting the forgery of the features that it enjoys.

show abstract

Section: Studies That Depend On Traditional Methodsmentioning

confidence: 99%

Section: Studies That Depend On Deep Learningmentioning

confidence: 99%

Section: Studies That Depend On Deep Learningmentioning

confidence: 99%

Section: Strengths and Weaknesses Of Previous Algorithmsmentioning

confidence: 99%

See 2 more Smart Citations

A survey on tamper detection techniques for digital images

Taha¹,

Ebraheem²

2022

AL-Rafidain Journal of Computer Sciences and Mathematics

View full text Add to dashboard Cite

show abstract

“…The classification accuracy can be used to assess the performance of the proposed method. The following formula can be used to determine the accuracy [29]:…”

Section: Performance Evaluation Measuresmentioning

confidence: 99%

Real and Altered Fingerprint Classification Based on Various Features and燙lassifiers

Hameed¹,

Ahmed²,

Okashi³

2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

Biometric recognition refers to the identification of individuals through their unique behavioral features (e.g., fingerprint, face, and iris). We need distinguishing characteristics to identify people, such as fingerprints, which are world-renowned as the most reliable method to identify people. The recognition of fingerprints has become a standard procedure in forensics, and different techniques are available for this purpose. Most current techniques lack interest in image enhancement and rely on high-dimensional features to generate classification models. Therefore, we proposed an effective fingerprint classification method for classifying the fingerprint image as authentic or altered since criminals and hackers routinely change their fingerprints to generate fake ones. In order to improve fingerprint classification accuracy, our proposed method used the most effective texture features and classifiers. Discriminant Analysis (DCA) and Gaussian Discriminant Analysis (GDA) are employed as classifiers, along with Histogram of Oriented Gradient (HOG) and Segmentation-based Feature Texture Analysis (SFTA) feature vectors as inputs. The performance of the classifiers is determined by assessing a range of feature sets, and the most accurate results are obtained. The proposed method is tested using a Sokoto Coventry Fingerprint Dataset (SOCOFing). The SOCOFing project includes 6,000 fingerprint images collected from 600 African people whose fingerprints were taken ten times. Three distinct degrees of obliteration, central rotation, and z-cut have been performed to obtain synthetically altered replicas of the genuine fingerprints. The proposal achieved massive success with a classification accuracy reaching 99%. The experimental results indicate that the proposed method for fingerprint classification is feasible and effective. The experiments also showed that the proposed SFTA-based GDA method outperformed state-of-art approaches in feature dimension and classification accuracy.

show abstract

Performance Analysis of Textured Contact Lens IRIS Detection Based on Manual Feature Engineering

Mahmood,

Ahmed

2024

Lecture Notes on Data Engineering and Communications Technologies

View full text Add to dashboard Cite

Effective Deep Features for Image Splicing Detection

Cited by 15 publications

References 19 publications

A survey on tamper detection techniques for digital images

A survey on tamper detection techniques for digital images

Real and Altered Fingerprint Classification Based on Various Features and燙lassifiers

Performance Analysis of Textured Contact Lens IRIS Detection Based on Manual Feature Engineering

Contact Info

Product

Resources

About