Breast density classification for cancer detection using DCT-PCA feature extraction and classifier ensemble
- Haque, Md Sarwar, Hassan, Md Rafiul, BinMakhashen, Galal, Owaidh, Abdullah, Kamruzzaman, Joarder
- Authors: Haque, Md Sarwar , Hassan, Md Rafiul , BinMakhashen, Galal , Owaidh, Abdullah , Kamruzzaman, Joarder
- Date: 2018
- Type: Text , Conference proceedings
- Relation: 17th International Conference on Intelligent Systems Design and Applications, ISDA 2017; Delhi, India; 14th-16th December 2017; published in Intelligent Systems Design and Applications (part of the Advances in Intelligent Systems and Computing book series) Vol. 736, p. 702-711
- Full Text:
- Reviewed:
- Description: It is well known that breast density in mammograms may hinder the accuracy of diagnosis of breast cancer. Although the dense breasts should be processed in a special manner, most of the research has treated dense breast almost the same as fatty. Consequently, the dense tissues in the breast are diagnosed as a developed cancer. In contrast, dense-fatty should be clearly distinguished before the diagnosis of cancerous or not cancerous breast. In this paper, we develop such a system that will automatically analyze mammograms and identify significant features. For feature extraction, we develop a novel system by combining a two-dimensional discrete cosine transform (2D-DCT) and a principal component analysis (PCA) to extract a minimal feature set of mammograms to differentiate breast density. These features are fed to three classifiers: Backpropagation Multilayer Perceptron (MLP), Support Vector Machine (SVM) and K Nearest Neighbour (KNN). A majority voting on the outputs of different machine learning tools is also investigated to enhance the classification performance. The results show that features extracted using a combination of DCT-PCA provide a very high classification performance while using a majority voting of classifiers outputs from MLP, SVM, and KNN.
- Authors: Haque, Md Sarwar , Hassan, Md Rafiul , BinMakhashen, Galal , Owaidh, Abdullah , Kamruzzaman, Joarder
- Date: 2018
- Type: Text , Conference proceedings
- Relation: 17th International Conference on Intelligent Systems Design and Applications, ISDA 2017; Delhi, India; 14th-16th December 2017; published in Intelligent Systems Design and Applications (part of the Advances in Intelligent Systems and Computing book series) Vol. 736, p. 702-711
- Full Text:
- Reviewed:
- Description: It is well known that breast density in mammograms may hinder the accuracy of diagnosis of breast cancer. Although the dense breasts should be processed in a special manner, most of the research has treated dense breast almost the same as fatty. Consequently, the dense tissues in the breast are diagnosed as a developed cancer. In contrast, dense-fatty should be clearly distinguished before the diagnosis of cancerous or not cancerous breast. In this paper, we develop such a system that will automatically analyze mammograms and identify significant features. For feature extraction, we develop a novel system by combining a two-dimensional discrete cosine transform (2D-DCT) and a principal component analysis (PCA) to extract a minimal feature set of mammograms to differentiate breast density. These features are fed to three classifiers: Backpropagation Multilayer Perceptron (MLP), Support Vector Machine (SVM) and K Nearest Neighbour (KNN). A majority voting on the outputs of different machine learning tools is also investigated to enhance the classification performance. The results show that features extracted using a combination of DCT-PCA provide a very high classification performance while using a majority voting of classifiers outputs from MLP, SVM, and KNN.
Passive detection of splicing and copy-move attacks in image forgery
- Islam, Mohammad, Kamruzzaman, Joarder, Karmakar, Gour, Murshed, Manzur, Kahandawa, Gayan
- Authors: Islam, Mohammad , Kamruzzaman, Joarder , Karmakar, Gour , Murshed, Manzur , Kahandawa, Gayan
- Date: 2018
- Type: Text , Conference proceedings , Conference paper
- Relation: 25th International Conference on Neural Information Processing, ICONIP 2018; Siem Reap, Cambodia; 13th-16th December 2018; published in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Vol. 11304 LNCS, p. 555-567
- Full Text:
- Reviewed:
- Description: Internet of Things (IoT) image sensors for surveillance and monitoring, digital cameras, smart phones and social media generate huge volume of digital images every day. Image splicing and copy-move attacks are the most common types of image forgery that can be done very easily using modern photo editing software. Recently, digital forensics has drawn much attention to detect such tampering on images. In this paper, we introduce a novel feature extraction technique, namely Sum of Relevant Inter-Cell Values (SRIV) using which we propose a passive (blind) image forgery detection method based on Discrete Cosine Transformation (DCT) and Local Binary Pattern (LBP). First, the input image is divided into non-overlapping blocks and 2D block DCT is applied to capture the changes of a tampered image in the frequency domain. Then LBP operator is applied to enhance the local changes among the neighbouring DCT coefficients, magnifying the changes in high frequency components resulting from splicing and copy-move attacks. The resulting LBP image is again divided into non-overlapping blocks. Finally, SRIV is applied on the LBP image blocks to extract features which are then fed into a Support Vector Machine (SVM) classifier to identify forged images from authentic ones. Extensive experiment on four well-known benchmark datasets of tampered images reveal the superiority of our method over recent state-of-the-art methods.
- Authors: Islam, Mohammad , Kamruzzaman, Joarder , Karmakar, Gour , Murshed, Manzur , Kahandawa, Gayan
- Date: 2018
- Type: Text , Conference proceedings , Conference paper
- Relation: 25th International Conference on Neural Information Processing, ICONIP 2018; Siem Reap, Cambodia; 13th-16th December 2018; published in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Vol. 11304 LNCS, p. 555-567
- Full Text:
- Reviewed:
- Description: Internet of Things (IoT) image sensors for surveillance and monitoring, digital cameras, smart phones and social media generate huge volume of digital images every day. Image splicing and copy-move attacks are the most common types of image forgery that can be done very easily using modern photo editing software. Recently, digital forensics has drawn much attention to detect such tampering on images. In this paper, we introduce a novel feature extraction technique, namely Sum of Relevant Inter-Cell Values (SRIV) using which we propose a passive (blind) image forgery detection method based on Discrete Cosine Transformation (DCT) and Local Binary Pattern (LBP). First, the input image is divided into non-overlapping blocks and 2D block DCT is applied to capture the changes of a tampered image in the frequency domain. Then LBP operator is applied to enhance the local changes among the neighbouring DCT coefficients, magnifying the changes in high frequency components resulting from splicing and copy-move attacks. The resulting LBP image is again divided into non-overlapping blocks. Finally, SRIV is applied on the LBP image blocks to extract features which are then fed into a Support Vector Machine (SVM) classifier to identify forged images from authentic ones. Extensive experiment on four well-known benchmark datasets of tampered images reveal the superiority of our method over recent state-of-the-art methods.
- «
- ‹
- 1
- ›
- »