Building a system to identify individuals through their speech recording can find its application in diverse areas, such as telephone shopping, voice mail and security control. However, building such systems is a tricky task because of the vast range of differences in the human voice. Thus, selecting strong features becomes very crucial for the recognition system. Therefore, a speaker recognition system based on new spin-image descriptors (SISR) is proposed in this paper. In the proposed system, circular windows (spins) are extracted from the frequency domain of the spectrogram image of the sound, and then a run length matrix is built for each spin, to work as a base for feature extraction tasks. Five different descriptors are generated from the run length matrix within each spin and the final feature vector is then used to populate a deep belief network for classification purpose. The proposed SISR system is evaluated using the English language Speech Database for Speaker Recognition (ELSDSR) database. The experimental results were achieved with 96.46 accuracy; showing that the proposed SISR system outperforms those reported in the related current research work in terms of recognition accuracy.
A new technique for embedding image data into another BMP image data is presented. The image data to be embedded is referred to as signature image, while the image into which the signature image is embedded is referred as host image. The host and the signature images are first partitioned into 8x8 blocks, discrete cosine transformed “DCT”, only significant coefficients are retained, the retained coefficients then inserted in the transformed block in a forward and backward zigzag scan direction. The result then inversely transformed and presented as a BMP image file. The peak signal-to-noise ratio (PSNR) is exploited to evaluate the objective visual quality of the host image compared with the original image.
One of the biomedical image problems is the appearance of the bubbles in the slide that could occur when air passes through the slide during the preparation process. These bubbles may complicate the process of analysing the histopathological images. The objective of this study is to remove the bubble noise from the histopathology images, and then predict the tissues that underlie it using the fuzzy controller in cases of remote pathological diagnosis. Fuzzy logic uses the linguistic definition to recognize the relationship between the input and the activity, rather than using difficult numerical equation. Mainly there are five parts, starting with accepting the image, passing through removing the bubbles, and ending with predict the tissues
... Show MoreInformation hiding strategies have recently gained popularity in a variety of fields. Digital audio, video, and images are increasingly being labelled with distinct but undetectable marks that may contain a hidden copyright notice or serial number, or even directly help to prevent unauthorized duplication. This approach is extended to medical images by hiding secret information in them using the structure of a different file format. The hidden information may be related to the patient. In this paper, a method for hiding secret information in DICOM images is proposed based on Discrete Wavelet Transform (DWT). Firstly. segmented all slices of a 3D-image into a specific block size and collecting the host image depend on a generated key
... Show MoreIn this paper, an algorithm for reconstruction of a completely lost blocks using Modified
Hybrid Transform. The algorithms examined in this paper do not require a DC estimation
method or interpolation. The reconstruction achieved using matrix manipulation based on
Modified Hybrid transform. Also adopted in this paper smart matrix (Detection Matrix) to detect
the missing blocks for the purpose of rebuilding it. We further asses the performance of the
Modified Hybrid Transform in lost block reconstruction application. Also this paper discusses
the effect of using multiwavelet and 3D Radon in lost block reconstruction.
Compression for color image is now necessary for transmission and storage in the data bases since the color gives a pleasing nature and natural for any object, so three composite techniques based color image compression is implemented to achieve image with high compression, no loss in original image, better performance and good image quality. These techniques are composite stationary wavelet technique (S), composite wavelet technique (W) and composite multi-wavelet technique (M). For the high energy sub-band of the 3rd level of each composite transform in each composite technique, the compression parameters are calculated. The best composite transform among the 27 types is the three levels of multi-wavelet
... Show More