Image enhancement is an important topic in the field of machine vision and image processing. Complex underwater environment poses great challenge for machine vision due to the turbid water medium. Serious scattering and absorption make the underwater images have complicated noise distribution. This paper proposes an underwater image enhancement model based on Encoding-Decoding deep CNN networks. We employ the convolution layers as encoding while deconvolution layers as decoding. The model achieves the image enhancement in an end-to-end adaptive way rather than considering the physical environment. We provide several comparison experiments with different datasets. Experiments show that it outperforms state-of-the art underwater image restoration methods in underwater image defogging, denoising and color enhancement. Finally, the model is employed to handle the underwater images with the different levels of noise and shows good performance.
7 November 2015 in Expert Opinion on Pharmacotherapy
Objectives: Controversy has been prompted based on drug interaction between proton pump inhibitors (PPIs) and aspirin/clopidogrel leading to weakened effects. However, whether such interaction was ...
24 October 2024 in IEEE Transactions on Aerospace and Electronic Systems
Pulse diversity phase coded (PDPC) waveform has received considerable attention recently because of its applications to obtain synthetic bandwidth, mitigate range-Doppler ambiguity or suppress artificial repeat jamming and clutter. However, when applied to Pulse-Doppler (PD) radar, high range-Doppler sidelobes will occur at the receiver, which cannot be improved by classical ambiguity function (AF) shaping methods. To address this problem, we first derive the discrete periodic ambiguity function (PAF) to characterize the range-Doppler processing of the PD radar. On this basis, we formulate a joint design problem for PDPC waveform and mismatched filter bank over the range-Doppler bins of interest under signal-to-noise ratio (SNR) loss and constant modulus constraints, which can be used to simultaneously suppress strong sidelobe interference and artificial repeat jamming. To solve the PAF shaping problem, which contains high-order polynomial (HOP) constraints, we generalize and apply the maximum block improvement type algorithm with the help of the alternating direction method of multipliers (ADMM) and introducing auxiliary variables. In particular, a time-varying step size strategy based on the dual theory is presented to enhance the convergence performance and refine the optimal solution. Several simulations are performed to show the effectiveness of the presented approach in improving the detection performance of PD radar, especially under low SNR scenarios.
17 November 2025 in IEEE Transactions on Aerospace and Electronic Systems
Interrupted sampling repeater jamming (ISRJ) can severely degrade the detection capability of linear frequency modulation (LFM) radars, due to its fast response and flexible modulation modes. To address the issue, this paper proposes a random inter-pulse frequency joint intra-pulse phase coding LFM waveform (RIF-PCLFM) and the corresponding method against ISRJ. The analytical expression of the ambiguity function (AF) is firstly derived. Based on this results, an excellent characteristic is revealed: the RIF-PCLFM signal inherits the high range resolution of the LFM signal while canceling the range-Doppler (RD) coupling effect. Thereby, we impose favorable constraints on the false targets generated by ISRJ, resulting in only a sparse deception effect rather than expansion or migration in the RD domain. This enables us to design a multi-channel one-shot phase compensation method for reconstructing, identifying, and suppressing false targets in the RD domain. Furthermore, to address the coherent processing challenge under inter-pulse phase nonlinearity, we propose an efficient non-uniform sparse Fourier transform (ENUSFT) algorithm with a decoupled inner-loop structure, which can well meet the real-time requirements of ISRJ suppression process. Multiple simulation results demonstrate that the proposed method exhibits outstanding detection and anti-jamming performance across various scenarios, with a processing efficiency reaching the millisecond level.