“…Compared with the methods presented in Refs. 19–23, the suggested method performs better in peak signal-to-noise ratio (PSNR). Even though the technique is robust, there has been no detailed investigation of overall embedding and recovery costs.…”
Section: State-of-the-art Watermarking In Deep-learning Environmentsmentioning
Recently, the demand for the generation, sharing, and storage of massive amounts of multimedia information-especially in the form of images-from different intelligent devices and sensors has increased drastically. This introduces issues including the illegal access and fraudulent usage of this information as well as other security concerns. Watermarking consists of embedding a watermark design in a digital cover and then later extracting it to provide a solution for ownership conflict and copyright violation issues involving the media data. Presently, in watermarking, the use of deep-learning approaches is incredibly beneficial due to their accuracy, superior results and strong learning ability. We present a comprehensive review of watermarking techniques in deep-learning environments. We start with basic concepts of traditional and learning-based digital watermarking; we later review the popular deep-learning model-based digital watermarking methods; then, we summarize and compare the most recent contribution in the literature; finally, we highlight obfuscation challenges and further research directions.
“…Compared with the methods presented in Refs. 19–23, the suggested method performs better in peak signal-to-noise ratio (PSNR). Even though the technique is robust, there has been no detailed investigation of overall embedding and recovery costs.…”
Section: State-of-the-art Watermarking In Deep-learning Environmentsmentioning
Recently, the demand for the generation, sharing, and storage of massive amounts of multimedia information-especially in the form of images-from different intelligent devices and sensors has increased drastically. This introduces issues including the illegal access and fraudulent usage of this information as well as other security concerns. Watermarking consists of embedding a watermark design in a digital cover and then later extracting it to provide a solution for ownership conflict and copyright violation issues involving the media data. Presently, in watermarking, the use of deep-learning approaches is incredibly beneficial due to their accuracy, superior results and strong learning ability. We present a comprehensive review of watermarking techniques in deep-learning environments. We start with basic concepts of traditional and learning-based digital watermarking; we later review the popular deep-learning model-based digital watermarking methods; then, we summarize and compare the most recent contribution in the literature; finally, we highlight obfuscation challenges and further research directions.
“…The wavelet entropy value can be used to find the small and short anomalies in the signal, and the sparse degree of the wavelet transform matrix can be used to suppress the irrelevant components to achieve effective signal extraction and eliminate the magnetism method. In 2011, Zhang Jian applied wavelet transform to detect magnetic anomalies [78]. According to the actual characteristics of magnetic abnormal signals, wavelet transform was adopted to process target signals polluted by non-Gaussian noise.…”
The geomagnetic field is the main magnetic field on the surface of the Earth, and its value is generally much larger than that of ferromagnetic objects. The existence of a geomagnetic field makes the ferromagnetic material magnetized, and the magnetized field will make the local total magnetic field abnormal, so it is called an anomalous magnetic field. This unusual magnetic field is a necessary condition for conducting magnetic anomaly detection (MAD). MAD is a widely used passive method for magnetic target detection, and its applications include surface ship target detection, the monitoring of underwater moving targets, land target detection and the identification of seismic activity for metal mining. MAD technology uses a high-sensitivity magnetometer to measure the target magnetic field. The magnetic field data are used to calculate the position, velocity, volume and other parameters of the target to identify and localize the ferromagnetic target. It is of great significance to study MAD data based on geomagnetic background. This paper reviews the MAD methods proposed by researchers in recent years and summarizes them into two categories. One is target based, and the other is noise based. The target-based group of detection methods involves typical magnetic search systems based on the assumption that the magnetometer and the target move relative to each other, which applies to the case where the target motion obeys a specific tracking time mode. The noise-based detection methods are based on statistical analyses of magnetometer noise and are suitable for situations in which assumptions about the mutual motion of the target and the magnetometer cannot be made. The magnetic dipole model is introduced in the second part of the paper, and then an algorithm based on the standard orthogonal basis function (OBF) decomposition is proposed. The algorithm parallels the target to a magnetic dipole and decomposes it into a linear combination of several standard OBFs. Solving for the coefficients of the basis function yields the signal energy function in the basis function space. The results show that the signal-to-noise ratio of the data processed by the OBF algorithm is significantly improved. The OBF can be further optimized; for example, when using a single magnetometer to conduct MAD, the five OBFs can be simplified to three OBFs; to locate the target more accurately when using two magnetometers to form the gradient magnetometer, the five OBFs can be simplified into four OBFs. The OBF algorithm is not very effective in the detection of non-Gaussian white noise, so
“…Li et al [10] proposed a watermarking algorithm based on the combination of redistributed discrete wavelet transform (RI-DWT) and singular value decomposition, which effectively improved the defects of poor robustness of DWT and IWT against geometric attacks. On this basis, Wei et al [11] proposed the redistributed invariant integer wavelet transform and applied it in the watermarking algorithm. The simulation results show that it not only has great advantages in resisting rotation attacks from different angles, but also has good performance in resisting filtering and noise.…”
In this paper, a blind watermarking algorithm based on redistributed invariant integer wavelet transform (RI-IWT) and BP network is proposed. The host image is processed by RI-DWT and QR decomposition, and the watermark is embedded in the low frequency and high frequency sub-band of the host image, so as to increase the embedding capacity. Moreover, we realize the blind extraction of watermark through BP network. In the extraction stage, the host image information and part of the watermark information are no longer required, thus improving the security of the algorithm. Finally, we simulate the algorithm and compare it with the relative algorithms. Simulation results show that the proposed algorithm has high invisibility and good robustness against JPEG attacks, rotation attacks, noise attacks.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.