针对红外图像增强过程中容易饱和、细节丢失等问题,提出一种参数自设定的双直方图均衡化方法。根据灰度级累积概率密度黄金比例值将原始图像划分为两个独立的子图像。结合原始图像曝光度和子图像灰度级区间信息,对每个子图像的直方图进...针对红外图像增强过程中容易饱和、细节丢失等问题,提出一种参数自设定的双直方图均衡化方法。根据灰度级累积概率密度黄金比例值将原始图像划分为两个独立的子图像。结合原始图像曝光度和子图像灰度级区间信息,对每个子图像的直方图进行多尺度自适应加权校正。基于校正后的直方图,对每个子图像分别作均衡化映射变换,最后合并子图像获得增强图像。在红外图像公开数据集INFRARED100上进行的测试显示,与亮度保持双直方图均衡化(Brightness Preserving Bi-Histogram Equalization,BBHE)、带平台限制的双直方图均衡化(Bi-histogram Equalization with a Plateau Limit,BHEPL)、基于曝光度的双直方图均衡化(Exposure based Sub-image Histogram Equalization,ESIHE)方法相比,所提方法增强的图像具有合适的平均对比度和更大的平均信息熵,在峰值信噪比(Peak Signal-to-Noise Ratio,PSNR)、结构相似度(Structural Similarity,SSIM)、绝对平均亮度偏差(Absolute Mean Brightness Error,AMBE)指标上平均提升至少17.2%、4.0%、56.2%。实验结果表明,所提方法对不同亮度特征的红外图像都有良好的适应性,可有效增强红外图像对象和背景之间的对比度,在噪声抑制、亮度和细节保持等方面优于同类方法。展开更多
This work proposes an improved inertia weight update method and position update method in Particle Swarm Optimization (PSO) to enhance the convergence and mean square error of channel equalizer. The search abilities o...This work proposes an improved inertia weight update method and position update method in Particle Swarm Optimization (PSO) to enhance the convergence and mean square error of channel equalizer. The search abilities of PSO are managed by the key parameter Inertia Weight (IW). A higher value leads to global search whereas a smaller value shifts the search to local which makes convergence faster. Different approaches are reported in literature to improve PSO by modifying inertia weight. This work investigates the performance of the existing PSO variants related to time varying inertia weight methods and proposes new strategies to improve the convergence and mean square error of channel equalizer. Also the position update method in PSO is modified to achieve better convergence in channel equalization. The simulation presents the enhanced performance of the proposed techniques in transversal and decision feedback models. The simulation results also analyze the superiority in linear and nonlinear channel conditions.展开更多
文摘针对红外图像增强过程中容易饱和、细节丢失等问题,提出一种参数自设定的双直方图均衡化方法。根据灰度级累积概率密度黄金比例值将原始图像划分为两个独立的子图像。结合原始图像曝光度和子图像灰度级区间信息,对每个子图像的直方图进行多尺度自适应加权校正。基于校正后的直方图,对每个子图像分别作均衡化映射变换,最后合并子图像获得增强图像。在红外图像公开数据集INFRARED100上进行的测试显示,与亮度保持双直方图均衡化(Brightness Preserving Bi-Histogram Equalization,BBHE)、带平台限制的双直方图均衡化(Bi-histogram Equalization with a Plateau Limit,BHEPL)、基于曝光度的双直方图均衡化(Exposure based Sub-image Histogram Equalization,ESIHE)方法相比,所提方法增强的图像具有合适的平均对比度和更大的平均信息熵,在峰值信噪比(Peak Signal-to-Noise Ratio,PSNR)、结构相似度(Structural Similarity,SSIM)、绝对平均亮度偏差(Absolute Mean Brightness Error,AMBE)指标上平均提升至少17.2%、4.0%、56.2%。实验结果表明,所提方法对不同亮度特征的红外图像都有良好的适应性,可有效增强红外图像对象和背景之间的对比度,在噪声抑制、亮度和细节保持等方面优于同类方法。
文摘This work proposes an improved inertia weight update method and position update method in Particle Swarm Optimization (PSO) to enhance the convergence and mean square error of channel equalizer. The search abilities of PSO are managed by the key parameter Inertia Weight (IW). A higher value leads to global search whereas a smaller value shifts the search to local which makes convergence faster. Different approaches are reported in literature to improve PSO by modifying inertia weight. This work investigates the performance of the existing PSO variants related to time varying inertia weight methods and proposes new strategies to improve the convergence and mean square error of channel equalizer. Also the position update method in PSO is modified to achieve better convergence in channel equalization. The simulation presents the enhanced performance of the proposed techniques in transversal and decision feedback models. The simulation results also analyze the superiority in linear and nonlinear channel conditions.
