General Background: Deep image matting is a fundamental task in computer vision, enabling precise foreground extraction from complex backgrounds, with applications in augmented reality, computer graphics, and video processing. Specific Background: Despite advancements in deep learning-based methods, preserving fine details such as hair and transparency remains a challenge. Knowledge Gap: Existing approaches struggle with accuracy and efficiency, necessitating novel techniques to enhance matting precision. Aims: This study integrates deep learning with fusion techniques to improve alpha matte estimation, proposing a lightweight U-Net model incorporating color-space fusion and preprocessing. Results: Experiments using the AdobeComposition-1k dataset demonstrate superior performance compared to traditional methods, achieving higher accuracy, faster processing speed, and improved boundary preservation. Novelty: The proposed model effectively combines deep learning with fusion techniques, enhancing matting quality while maintaining robustness across various environmental conditions. Implications: These findings highlight the potential of integrating fusion techniques with deep learning for image matting, offering valuable insights for future research in automated image processing applications, including augmented reality, gaming, and interactive video technologies. Highlights: Better Precision: Fusion techniques enhance fine detail preservation. Faster Processing: Lightweight U-Net improves speed and accuracy. Wide Applications: Useful for AR, gaming, and video processing. Keywords: Deep image matting, computer vision, deep learning, fusion techniques, U-Net
Huge number of medical images are generated and needs for more storage capacity and bandwidth for transferring over the networks. Hybrid DWT-DCT compression algorithm is applied to compress the medical images by exploiting the features of both techniques. Discrete Wavelet Transform (DWT) coding is applied to image YCbCr color model which decompose image bands into four subbands (LL, HL, LH and HH). The LL subband is transformed into low and high frequency components using Discrete Cosine Transform (DCT) to be quantize by scalar quantization that was applied on all image bands, the quantization parameters where reduced by half for the luminance band while it is the same for the chrominance bands to preserve the image quality, the zig
... Show MoreA novel encapsulated deep eutectic solvent (DES) was introduced for biodiesel production via a two-step process. The DES was encapsulated in medical capsules and were used to reduce the free fatty acid (FFA) content of acidic crude palm oil (ACPO) to the minimum acceptable level (< 1%). The DES was synthesized from methyltriphenylphosphonium bromide (MTPB) and p-toluenesulfonic acid (PTSA). The effects pertaining to different operating conditions such as capsule dosage, reaction time, molar ratio, and reaction temperature were optimized. The FFA content of ACPO was reduced from existing 9.61% to less than 1% under optimum operating conditions. This indicated that encapsulated MTPB-DES performed high catalytic activity in FFA esterificatio
... Show MoreA standard theoretical neutron energy flux distribution is achieved for the triton-triton nuclear fusion reaction in the range of triton energy about ≤10 MeV. This distribution give raises an evidence to provide the global calculations including the characteristics fusion parameters governing the T-T fusion reaction.
In this study, an efficient compression system is introduced, it is based on using wavelet transform and two types of 3Dimension (3D) surface representations (i.e., Cubic Bezier Interpolation (CBI)) and 1 st order polynomial approximation. Each one is applied on different scales of the image; CBI is applied on the wide area of the image in order to prune the image components that show large scale variation, while the 1 st order polynomial is applied on the small area of residue component (i.e., after subtracting the cubic Bezier from the image) in order to prune the local smoothing components and getting better compression gain. Then, the produced cubic Bezier surface is subtracted from the image signal to get the residue component. Then, t
... Show MoreAmputation of the upper limb significantly hinders the ability of patients to perform activities of daily living. To address this challenge, this paper introduces a novel approach that combines non-invasive methods, specifically Electroencephalography (EEG) and Electromyography (EMG) signals, with advanced machine learning techniques to recognize upper limb movements. The objective is to improve the control and functionality of prosthetic upper limbs through effective pattern recognition. The proposed methodology involves the fusion of EMG and EEG signals, which are processed using time-frequency domain feature extraction techniques. This enables the classification of seven distinct hand and wrist movements. The experiments conducte
... Show MoreThere are many images you need to large Khoznah space With the continued evolution of storage technology for computers, there is a need nailed required to reduce Alkhoznip space for pictures and image compression in a good way, the conversion method Alamueja
Many image processing and machine learning applications require sufficient image feature selection and representation. This can be achieved by imitating human ability to process visual information. One such ability is that human eyes are much more sensitive to changes in the intensity (luminance) than the color information. In this paper, we present how to exploit luminance information, organized in a pyramid structure, to transfer properties between two images. Two applications are presented to demonstrate the results of using luminance channel in the similarity metric of two images. These are image generation; where a target image is to be generated from a source one, and image colorization; where color information is to be browsed from o
... Show MoreIn this work, the fusion cross section , fusion barrier distribution and the probability of fusion have been investigated by coupled channel method for the systems 46Ti+64Ni, 40Ca+194Pt and 40Ar+148Sm with semi-classical and quantum mechanical approach using SCF and CCFULL Fortran codes respectively. The results for these calculations are compared with available experimental data. The results show that the quantum calculations agree better with experimental data, especially bellow the Coulomb barrier, for the studied systems while above this barrier, the two codes reproduce the data.