二维条码作为一种能够存储大量信息的图形标识,在工业生产的自动化控制、物流管理、质量追溯以及运输的信息交换等环节起着非常重要的作用。二维条码的高精度识别是实现快速、准确信息交换的基础。但是受拍摄环境和拍摄设备精度的制约,...二维条码作为一种能够存储大量信息的图形标识,在工业生产的自动化控制、物流管理、质量追溯以及运输的信息交换等环节起着非常重要的作用。二维条码的高精度识别是实现快速、准确信息交换的基础。但是受拍摄环境和拍摄设备精度的制约,经常由于分辨率低而无法正确识别。针对此问题,提出了一种面向实际降质二维条码的超分辨率重建算法。考虑到实际降质的复杂性,提出了基于降质先验的超分辨率算法。首先,设计了一个降质先验信息编码器,用于提取和编码因拍摄环境和设备限制导致的图像质量降低的相关信息。然后,提出了一个降质先验引导模块,使用编码器提取的信息来引导主体结构的特征重建,包括降质特征图引导与降质先验引导两部分。由于目前缺少相关数据集,所以率先构建了真实退化条件下的二维条码超分辨率数据集(包含4944对低分辨率-高分辨率二维条码图像)。考虑到真实数据对之间有轻微位移,引入位移不敏感的损失函数对网络进行优化。实验表明,所提方法重建结果的峰值信噪比(Peak Signal to Noise Ratio,PSNR)、结构相似性(Structural Similarity,SSIM)以及扫出率这3种指标均优于5种经典的超分辨率重建算法,充分说明了所提方法的优越性。展开更多
Underwater vehicles have already adopted self-correcting directional guidance algorithms based on multi-beam self-guidance systems, not waiting for research to determine the most effective algorithms. The main challen...Underwater vehicles have already adopted self-correcting directional guidance algorithms based on multi-beam self-guidance systems, not waiting for research to determine the most effective algorithms. The main challenges facing research on these guidance systems have been effective modeling of the guidance algorithm and a means to analyze the simulation results. A simulation structure based on Simulink that dealt with both issues was proposed. Initially, a mathematical model of relative motion between the vehicle and the target was developed, which was then encapsulated as a subsystem. Next, steps for constructing a model of the self-correcting guidance algorithm based on the Stateflow module were examined in detail. Finally, a 3-D model of the vehicle and target was created in VRML, and by processing mathematical results, the model was shown moving in a visual environment. This process gives more intuitive results for analyzing the simulation. The results showed that the simulation structure performs well. The simulation program heavily used modularization and encapsulation, so has broad applicability to simulations of other dynamic systems.展开更多
文摘二维条码作为一种能够存储大量信息的图形标识,在工业生产的自动化控制、物流管理、质量追溯以及运输的信息交换等环节起着非常重要的作用。二维条码的高精度识别是实现快速、准确信息交换的基础。但是受拍摄环境和拍摄设备精度的制约,经常由于分辨率低而无法正确识别。针对此问题,提出了一种面向实际降质二维条码的超分辨率重建算法。考虑到实际降质的复杂性,提出了基于降质先验的超分辨率算法。首先,设计了一个降质先验信息编码器,用于提取和编码因拍摄环境和设备限制导致的图像质量降低的相关信息。然后,提出了一个降质先验引导模块,使用编码器提取的信息来引导主体结构的特征重建,包括降质特征图引导与降质先验引导两部分。由于目前缺少相关数据集,所以率先构建了真实退化条件下的二维条码超分辨率数据集(包含4944对低分辨率-高分辨率二维条码图像)。考虑到真实数据对之间有轻微位移,引入位移不敏感的损失函数对网络进行优化。实验表明,所提方法重建结果的峰值信噪比(Peak Signal to Noise Ratio,PSNR)、结构相似性(Structural Similarity,SSIM)以及扫出率这3种指标均优于5种经典的超分辨率重建算法,充分说明了所提方法的优越性。
文摘Underwater vehicles have already adopted self-correcting directional guidance algorithms based on multi-beam self-guidance systems, not waiting for research to determine the most effective algorithms. The main challenges facing research on these guidance systems have been effective modeling of the guidance algorithm and a means to analyze the simulation results. A simulation structure based on Simulink that dealt with both issues was proposed. Initially, a mathematical model of relative motion between the vehicle and the target was developed, which was then encapsulated as a subsystem. Next, steps for constructing a model of the self-correcting guidance algorithm based on the Stateflow module were examined in detail. Finally, a 3-D model of the vehicle and target was created in VRML, and by processing mathematical results, the model was shown moving in a visual environment. This process gives more intuitive results for analyzing the simulation. The results showed that the simulation structure performs well. The simulation program heavily used modularization and encapsulation, so has broad applicability to simulations of other dynamic systems.
