The large-capacity on-chip cache often causes timing problems or erroneous transmission in the high-speed acquisition of massive data. However, many factors always affect communication performance system. In this paper, a data system (DAS) based on DMA double mechanism is introduced. Aurora protocol used for optical fiber transmission, and line rate dynamic adaptive design X1 channel realized. SDRAM controller, dual targeted optimized, including memory ping-pong operation arbiter. addition,...

Salient ship detection plays an important role in ensuring the safety of maritime transportation and navigation. However, due to influence waves, special weather, illumination on sea, existing saliency methods are still unable achieve effective a complex marine environment. To solve problem, this paper proposed novel method based attention nested U-Structure (AU2Net). First, make up for shortcomings U-shaped structure, pyramid pooling module (PPM) global guidance paths (GGPs) designed guide...

As a kind of marine vehicles, Unmanned Surface Vehicles (USV) are widely used in military and civilian fields because their low cost, good concealment, strong mobility high speed. High-precision detection obstacles plays an important role USV autonomous navigation, which ensures its subsequent path planning. In order to further improve obstacle performance, we propose encoder-decoder architecture named Fusion Refinement Network (FRN). The encoder part with deeper network structure enables it...

Visual ship detection is a crucial component in recognition systems for detecting objects at different scales. However, there remains challenge multiscale ships with the influence of various sizes and complex backgrounds. We propose an efficient multilayer attention receptive fusion network (MARN) that based on well-established YOLOv4. MARN one-stage algorithm essentially extracts semantic information from feature maps scales to highlight salient features ships, thereby improving performance...

Face detection has been an independent technology playing important role in more and fields, which makes it necessary urgent to have its architecture reconfigurable meet different demands on capabilities. This paper proposed a face architecture, could be adjusted by the user according background, sensor resolution, accuracy speed situations. adjustable mode reconfiguration simple efficient, is especially suitable for portable mobile terminals whose working condition often changes frequently....

Abstract Rain streaks could blur and distort images, significantly impacting further image processing. Single-image deraining is a hotspot has practical application value, while most existing methods still have problems such as residual rain inadequate recovery of detail textures. To address these issues, we propose Residual Contextual Hourglass Network (RCHNet) for single-image deraining, which adapt to remove in complex environments. Firstly, contextual distillation block presented obtain...

Existing shipborne AIS (Automatic Identification System) based on the text mode, larger amount of information, is not conducive to visual observation ships around, this paper presents a graphical interface for information display way. The adopts method system with embedded explain system's hardware architecture and interface, analyzes key software technology at same time gives result.

Detecting small infrared targets is a difficult but important task in highly cluttered coastal surveillance. The paper proposed method called low-rank and sparse decomposition based frame difference to improve the detection performance of surveillance system. First, used adjacent frames detect candidate object regions which we are most interested in. Then further exclude clutters by matrix recovery. Finally, extracted from recovered target component local self-adaptive threshold. experiment...

Face detection has been playing an important role in more and fields, which makes it necessary urgent to have its architecture reconfigurable satisfy different requirements on capabilities. The paper proposes a face architecture, could be adjusted by the users according environment, sensor resolution, of accuracy speed. This user adjustable mode reconfiguration simple efficient, especially benefits for application portable mobile terminals whose working condition often changes frequently....

As it is difficult to detect small ship targets from complex background in an IR image because of the impact kinds noises, and smallness target, this paper puts forward a method prediction utilizing neural network. The estimates input nonlinearly, gains residual error image, detects target. By comparison, NARX whose hidden layer transfer function L-M algorithm can achieve more excellent target detection than other networks like BP training speed stability, thus research valuable field

Based on the analysis of characteristics infrared ship target in complex background, a multi-frame image fusion method using morphological difference pyramid or wavelet transform was proposed to get more clear and reliable image. The experimental results show that both methods can overcome noise interfere background certain extent, which enhance capabilities detection situation awareness, but there are still some differences their work speed performance.