Optimization-driven steganographic system based on fused maps and blowfish encryption.

This paper presents a lightweight framework for secure metadata embedding in unmanned aerial vehicle (UAV) communication. An on-board convolutional neural network (CNN) classifier generates image predictions, which are directly embedded into the corresponding image using Least Significant Bit (LSB) steganography. This approach eliminates the need for separate metadata channels, thereby enhancing confidentiality and improving bandwidth efficiency. The proposed framework is validated on a subset of the CIFAR-10 benchmark dataset as well as the UAV-specific VisDrone2019-DET dataset. The predicted class labels and confidence scores are embedded within the images and can be extracted without error. The methodology includes CNN model design, LSB embedding and extraction, and evaluation across classification accuracy and image quality metrics. The evaluation employs accuracy, Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index Measure (SSIM). Experiments show classification accuracies of 33.0% on CIFAR- 10 and 48.36% on VisDrone. Visual quality is preserved, with PSNR values above 63.0 dB and SSIM scores near 1.0.The results confirm that the proposed method achieves secure metadata embedding with negligible visual distortion, offering an efficient and lightweight solution for UAV communication

This paper presents hiding one to eight gray images or texts (secret information) inside color image (cover image) based on Least Significant Bits (LSB) algorithm. The main idea is to insert the secret message (one to eight texts or one to eight gray images) in the three LSB of the cover image (color image). The color cover image RGB model (R channel, G channel and B channel) is separated into three cover images (the red cover, green cover, and blue cover). Each one of these three cover images is used to hide secret messages (gray image (two secret images in R channel, three secret images in G channel and three secret images in B channel) or text). The experiments and comparative studies show that the algorithms are characterized by many features of the ability of hiding huge data, and then the ability of extracting secret message without errors. Beside the reconstructed image, has efficacies (to human acquaintance) according to peak signal to noise ratio (PSNR) and mean square error (MSE), also retain both the explicitness and the characteristics of the both secret message and cover image.

In this study, we propose a CNN-GAN-based real-time processing technique for filtering images of underwater cables used in power systems. This addresses the excessive interference impurities that are frequently observed in images captured by remotely operated vehicles (ROVs). The process begins with the input of the original image into the convolutional neural network (CNN). Subsequently, the training outcomes, which serve as input parameters for the generative adversarial network (GAN), facilitate the filtering process. The system also calculates both the structural similarity index measure (SSIM) and peak signal-to-noise ratio (PSNR), performing model updates via backward propagation. This technique utilizes deep learning technologies to achieve rapid, real-time filtering of underwater cable images. The experimental results reveal that the loss function of the CNN reaches 0.16 with an accuracy of 97.5%, while the loss function of the adversarial GAN network approaches 0.05. Compared with traditional methods such as DDN, JORDOR, RESCAN, and PRENet, the proposed CNN-GAN algorithm exhibits superior performance, as evidenced by the higher PSNR and SSIM values. Specifically, for clear water images, the PSNR reaches 29.86 dB and the SSIM is 0.9045. For severely polluted images, the PSNR is 28.67 dB and the SSIM is 0.8965, while for unevenly illuminated images, the PSNR and SSIM values are 24.37 dB and 0.88, respectively. These enhancements significantly benefit the monitoring and maintenance of power systems.

Steganography is a technique for hiding secret data within digital media such as images, audio, and video without causing noticeable visual changes. In video media, this technique offers advantages because each frame can be utilized dynamically, resulting in a larger data embedding capacity. However, conventional methods such as fixed-number Least Significant Bit (LSB) embedding still face limitations in balancing visual quality, embedding capacity, and resistance to compression or noise. To address these challenges, this study proposes an Adaptive Video Steganography Method based on Multi-Bit LSB that employs brightness, texture, and motion analysis for each frame to determine the number of embedding bits adaptively. The system adjusts the embedding capacity according to the local characteristics of the video: areas with high texture or rapid motion are assigned more bits, while static or low-texture areas use fewer bits to preserve visual quality. After the embedding process, the video quality is evaluated using PSNR (Peak Signal-to-Noise Ratio) and SSIM (Structural Similarity Index Measurement) to assess its similarity to the original video. Experimental results show a PSNR value of 45.86 dB and an SSIM value of 0.9441, Thus, the proposed adaptive method proves to be efficient, robust against disturbances, and capable of maintaining data security without compromising visual quality, making it highly suitable for implementation in multimedia-based information security systems.

WITHDRAWN: Contrasting the performance metrics of discrete transformations on digital image steganography using artificial intelligence

Steganography is a process to conceal the existence of a message. Image Steganography involves hiding the message onto a cover image to obtain a resultant stego image. This paper discusses a procedure for hiding coloured message images onto coloured cover images using a lossy Image Steganography technique. The secret images undergo Discrete Cosine Transformation, quantization and AES encryption before embedding using the Least Significant Bit embedding (LSB) technique onto the cover image. Finally, Peak signal-to-noise ratio (PSNR), Structural Similarity Index measurement (SSIM), zero normalized cross-correlation (ZNCC) are used for comparing the cover and stego images as well as the extracted and original images.

The aim of this study is to remove scattered photons and beam hardening effect in cone beam CT (CBCT) images and make an image available for treatment planning. To remove scattered photons and beam hardening effect, a convolutional neural network (CNN) was used, and trained with distorted projection data including scattered photons and beam hardening effect and supervised projection data calculated with monochromatic X-rays. The number of training projection data was 17,280 with data augmentation and that of test projection data was 540. The performance of the CNN was investigated in terms of the number of photons in the projection data used in the training of the network. Projection data of pelvic CBCT images (32 cases) were calculated with a Monte Carlo simulation with six different count levels ranging from 0.5 to 3 million counts/pixel. For the evaluation of corrected images, the peak signal-to-noise ratio (PSNR), the structural similarity index measure (SSIM), and the sum of absolute difference (SAD) were used. The results of simulations showed that the CNN could effectively remove scattered photons and beam hardening effect, and the PSNR, the SSIM, and the SAD significantly improved. It was also found that the number of photons in the training projection data was important in correction accuracy. Furthermore, a CNN model trained with projection data with a sufficient number of photons could yield good performance even though a small number of photons were used in the input projection data.

This paper proposed to extend the ATD steganographic method to work with color images and conduct an experimental study of the modified algorithm. In the original ATD algorithm, which works on grayscale images, The main idea is embedding the secret message as the ternary number and in each pixel of cover image is embedding two ternary numbers. For color images. we propose embedding four ternary numbers within each pixel of the cover image. As is well known, color images are composed of three primary color channels: red, green, and blue. The final color is generated by blending specific intensities of these three channels. Accordingly, three embedding combinations (112,121,211) were proposed to determine how the four ternary numbers are distributed among the red, green, and blue channels of each pixel. Each such combination corresponds to approximately 8 BPP for example, using the combination 211 means embedding two ternary numbers in the red channel (R), and one ternary number in each of the green (G) and blue (B) channels. By using the Matlab program, we compared the results on eight color images selected as cover images to evaluate the performance of the proposed algorithm. Image quality metrics, such as Peak Signal-to-Noise Ratio (PSNR) and Mean Squared Error (MSE), were used to evaluate the algorithm’s performance according to our analysis, the value of PSNR is not affected by the scheme but depends on image; PSNR is in range from 34.50 dB to 37.57 dB in same embedding capacity 8 BPP. Moreover, SNR values across the different embedding combinations, suggesting comparable performance among them. Keywords: Steganography, ATD, Peak Signal to Noise Ratio (PSNR), Embedding Capacity, Bit Per Pixel (BPP).

Blurring facial images can significantly degrade the performance of face recognition and video surveillance applications. Therefore, an effective image restoration method is essential to address this issue. However, existing methods struggle with varying levels of blur severity, limiting their effectiveness. This study proposes a facial image restoration approach that integrates blur severity classification using a Convolutional Neural Network (CNN) with a U-Net-based deblurring model to overcome these challenges. This method ensures that each blurred image is processed using the most suiTable deblurring model, optimizing the restoration process. The dataset used in this study is Flickr-Faces-HQ (FFHQ), to which a Gaussian blur is applied and categorized into five levels: very low, low, medium, high, and very high. They employ the Peak Signal-to-Noise Ratio (PSNR) and the Structural Similarity Index Measure (SSIM) as quantitative metrics to evaluate the model's performance. Experimental results show that the proposed method consistently outperforms fixed kernel size and multi-kernel size deblurring approaches across all blur severities. Specifically, our method achieves a PSNR of 40.001 dB for very low blur severity and an SSIM of 0.990. For low severity, it attains a PSNR of 31.104 dB and an SSIM of 0.946. For medium severity, the results are 27.995 dB (PSNR) and 0.874 (SSIM). At high severity, our model achieves a PSNR of 28.896 dB and an SSIM of 0.855. Finally, for very high severity, the PSNR drops to 26.566 dB, with an SSIM of 0.812. These results demonstrate the effectiveness of the proposed method in enhancing image clarity across different blur levels while preserving facial details. This research contributes to the development of more adaptive and efficient image restoration techniques, particularly for applications that require high-quality facial images, such as face recognition and video surveillance.

Recent work has shown that deep convolutional neural network is capable of solving inverse problems in computational imaging, and recovering the stress field of the loaded object from the photoelastic fringe pattern can also be regarded as an inverse problem solving process. However, the formation of the fringe pattern is affected by the geometry of the specimen and experimental configuration. When the loaded object produces complex fringe distribution, the traditional stress analysis methods still face difficulty in unwrapping. In this study, a deep convolutional neural network based on the encoder–decoder structure is proposed, which can accurately decode stress distribution information from complex photoelastic fringe images generated under different experimental configurations. The proposed method is validated on a synthetic dataset, and the quality of stress distribution images generated by the network model is evaluated using mean squared error (MSE), structural similarity index measure (SSIM), peak signal-to-noise ratio (PSNR), and other evaluation indexes. The results show that the proposed stress recovery network can achieve an average performance of more than 0.99 on the SSIM.

Steganography is a method of hiding sensitive information within an ordinary file to avoid detection. When the file is received, the hidden information is extracted from the original file. Various types of files can be used to conceal the information, including images, audio files, video files, and text files. This research paper focuses on the use of image files as the medium. To create a stego-image, a secret image is embedded into a cover image. Both the cover and secret images are subjected to a Daubechies 3 Discrete Wavelet Transformation (DWT), then encrypted using Arnold's Cat Map algorithm and fused using a simple alpha blending technique with the cover image. DWT enhances the invisibility of hidden information, making it more challenging to detect by visual inspection or statistical analysis. Arnold’s cat map adds an extra layer of security, making it more difficult for an attacker to extract the message. Alpha blending helps to maintain the visual quality and integrity of the cover image, reducing the chances of noticeable artifacts or anomalies that may arise due to the embedding process. The resulting stego-image appears visually like the original cover image, making it less suspicious to potential attackers. Experimental results have been made on multiple area such as the best value of alpha, which DWT family to use, and different combinations of images as cover and payload images. The results demonstrate that our proposed hybrid model achieves a high level of security, robustness, and imperceptibility in image steganography. Moreover, it outperforms existing methods in terms of image quality metrics, such as achieving a Peak Signal-to-Noise Ratio (PSNR) value of up to 70.8 and a Structural Similarity Index (SSIM) value of up to 0.98982. These metrics indicate that the stego-image retains a remarkable level of visual fidelity and similarity to the cover image, thereby enhancing the overall effectiveness of the steganography technique.

In this paper, an experimental analysis and study of the Modified Signed-Digit (MSD) method was conducted, which aims to hide a quantity of secret data represented in a grayscale image of size 512 × 512 in a set of cover images. Some quality metrics such as signal-to-noise ratio (PSNR), mean squared error (MSE), and embedding capacity (BPP) were measured. As well as calculating, the optimal number of cover images, the time taken in the hiding process, and the consumed memory. The experiments were conducted two different sizes of cover images, where the number of cover images required for the secret image is determined according to some inputs such as the number of pixels n in each group of the cover image and the number of bits in each block of the secret image L=n+1. Also, four different values of n, (2, 3, 4, 5). We tried to find the best value for the number of bits in each block L of the secret image and in each case of n to achieve the best case for image quality PSNR, MSE for both cover image sizes, and on the other hand, less time memory consumption. By using the Matlab program, we compared the results at different cover image sizes, we found that as the larger size used, the fewer cover images are needed and less time consumption. On the other hand, more memory consumption. For the other metrics, PSNR, MSE and BPP, close results were obtained. Finally, the best choice remains related to the user and the type of application or the required performance, whether in image quality or in less memory and time consumption with a larger embedding capacity. Keywords: MSD algorithm, Signal-to-noise ratio (PSNR), Mean square error (MSE), Bit-per-pixel (BPP), Memory consumption, Time consumption.

Objective. The study aims to systematically characterize the effect of CT parameter variations on images and lung radiomic and deep features, and to evaluate the ability of different image harmonization methods to mitigate the observed variations.Approach. A retrospective in-house sinogram dataset of 100 low-dose chest CT scans was reconstructed by varying radiation dose (100%, 25%, 10%) and reconstruction kernels (smooth, medium, sharp). A set of image processing, convolutional neural network (CNNs), and generative adversarial network-based (GANs) methods were trained to harmonize all image conditions to a reference condition (100% dose, medium kernel). Harmonized scans were evaluated for image similarity using peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and learned perceptual image patch similarity (LPIPS), and for the reproducibility of radiomic and deep features using concordance correlation coefficient (CCC).Main Results. CNNs consistently yielded higher image similarity metrics amongst others; for Sharp/10%, which exhibited the poorest visual similarity, PSNR increased from a mean ± CI of 17.763 ± 0.492 to 31.925 ± 0.571, SSIM from 0.219 ± 0.009 to 0.754 ± 0.017, and LPIPS decreased from 0.490 ± 0.005 to 0.275 ± 0.016. Texture-based radiomic features exhibited a greater degree of variability across conditions, i.e. a CCC of 0.500 ± 0.332, compared to intensity-based features (0.972 ± 0.045). GANs achieved the highest CCC (0.969 ± 0.009 for radiomic and 0.841 ± 0.070 for deep features) amongst others. CNNs are suitable if downstream applications necessitate visual interpretation of images, whereas GANs are better alternatives for generating reproducible quantitative image features needed for machine learning applications.Significance. Understanding the efficacy of harmonization in addressing multi-parameter variability is crucial for optimizing diagnostic accuracy and a critical step toward building generalizable models suitable for clinical use.

This study introduces a cutting-edge image encryption algorithm aimed at elevating security standards. The Pan-Tompkins Algorithm (PTA) for key generation is proposed for the first time in this study. Additionally, employing steganography through the Least Significant Bit (LSB) method for embedding keys within the encrypted image enhances secure key distribution, thereby fortifying the encryption process. On the other hand, the integration of advanced algorithms, such as Zigzag scanning, the Affine Image Encryption Algorithm (AA), and the Vigenere Image Encryption Algorithm (VA), constitutes the fundamental innovation of the proposed image encryption algorithm. The proposed algorithm is named PanAAVA:Affine Algorithm and Vigenere Algorithm Encryption with PTA-Based Key Generation. The PanAAVA algorithm ensures unparalleled security by encrypting the positions and values of pixels using AA and VA. Notably, using PTA for key generation marks a distinctive and new key generation method feature of the algorithm. To assess the effectiveness of the PanAAVA, a comprehensive comparative analysis is conducted against well-established encryption methodologies, including Lena, Baboon, Airplane, and Pepper.The PanAAVA demonstrates exceptional proficiency in histogram analysis. The PanAAVA demonstrates a Unified Average Changing Intensity (UACI) of 33.4044%. Additionally, the Number of Pixels Change Rate (NPCR) is measured at 99.7442%, showcasing the algorithm’s effectiveness in inducing significant pixel changes. The proposed algorithm’s Mean Square Error (MSE) is calculated at 3.20679E5%. The proposed algorithm’s Peak Signal to Noise Ratio (PSNR) is recorded at 9.512475. The Key Space Size of the proposed algorithm is measured at 2209. Regarding correlation analysis, the PanAAVA achieves a high correlation score of 7.9996. The proposed algorithm successfully passes the National Institute of Standards and Technology (NIST) analysis, demonstrating a remarkably strong correlation close to 0 and a Structural Similarity Index Measure (SSIM) of 0.9977. Furthermore, regarding quantum communication, the proposed algorithm maintains stable key rates of 47.5 ± 0.8 kHz during the day and 50.9 ± 0.7 kHz at night. Additionally, PanAAVA achieves low Quantum Bit Error Rate (QBER) values of 4.77 ± 0.02, ensuring reliable and secure communication. The PanAAVA also demonstrates robust asymmetries at 49.81 ± 0.02 and 50.14 ± 0.03 for a crystal length of 20 mm. highlighting PanAAVA’s adaptability and effectiveness in different scenarios. PanAAVA outperforms other encryption algorithms concerning performance measurements and comparisons. In conclusion, the PanAAVA emerges as a beacon of superior security capabilities and innovation in image encryption, showcasing the potential to redefine standards in the field.

Speckle noise in Intravascular Ultrasound (IVUS) images poses a significant challenge, necessitating early elimination of these artifacts. In this study, two novel methods were designed: a denoising convolutional autoencoder and a modified non-local means algorithm using particle swarm optimization (PSO) for automatic parameter selection. This research delved into assessing the efficacy of various denoising models, encompassing spatial domain techniques (median filter, non-local means, optimized non-local means), transform domain methods (wavelet, curvelet), and deep learning approaches (denoising autoencoder). A wide array of data augmentation techniques was employed during the training of the autoencoder model. Evaluation was conducted using two key metrics: Structural Similarity Index Measure (SSIM) and Peak Signal-to-Noise Ratio (PSNR). The findings indicate the robustness of the autoencoder model with the proposed augmentation strategy across different speckle noise variances, highlighting its superiority over traditional methods. Moreover, the study underscores the enhanced performance of modified non-local means with particle swarm optimization (PSO) compared to manual non-local means.