CAJ | 학술논문

在浓雾环境下进行视频拍摄会因大气粒子的散射作用导致图像成像效果极差,通过图像增强算法设计,提高视频图像的去雾速率和成像质量.当前对浓雾环境下拍摄出来的景物图像采用暗原色增强方法,对能见度和对比度较低的图像的增强效果不好.提出一种基于改进的多尺度加权平均Ret?inex算法的雾天视频图像增强算法.采用Gamma-Gamma分布的方法构建大气散射物理模型,利用简化大气传播原理,得到去雾后的图像的视频图像的暗原色特征,实现雾天环境下视频图像成像机理和增强模型总体设计.浓雾环境下提取多尺度Retinex图像暗原色特征,采用多尺度加权平均Retinex算法,对图像的灰度像素值进行自适应增强,指导雾化图像进行盒子滤波实现去雾处理,得到图像的亮度、对比度和结构相似度的增强结果.仿真表明,该算法具有较好的去雾性能和图像增强性能,提高了去雾速率和浓雾下的视频图像成像质量.
재농무배경하진행시빈박섭회인대기입자적산사작용도치도상성상효과겁차,통과도상증강산법설계,제고시빈도상적거무속솔화성상질량.당전대농무배경하박섭출래적경물도상채용암원색증강방법,대능견도화대비도교저적도상적증강효과불호.제출일충기우개진적다척도가권평균Ret?inex산법적무천시빈도상증강산법.채용Gamma-Gamma분포적방법구건대기산사물리모형,이용간화대기전파원리,득도거무후적도상적시빈도상적암원색특정,실현무천배경하시빈도상성상궤리화증강모형총체설계.농무배경하제취다척도Retinex도상암원색특정,채용다척도가권평균Retinex산법,대도상적회도상소치진행자괄응증강,지도무화도상진행합자려파실현거무처리,득도도상적량도、대비도화결구상사도적증강결과.방진표명,해산법구유교호적거무성능화도상증강성능,제고료거무속솔화농무하적시빈도상성상질량.
The video camera will range imaging effect caused by atmospheric scattering particles in fog environment, through the design of image enhancement algorithm, improve video defogging rate and imaging quality. At present on the scene image fog environment shooting out of the dark color enhancement method using, not good to enhance the effect of vis-ibility and a low contrast image. An enhanced algorithm of foggy video image is proposed based on multi-scale Retinex weighted average algorithm. The construction of the atmosphere scattering model with the method of Gamma-Gamma distri-bution, using a simplified atmospheric propagation theory, extracting multi-scale Retinex image dark channel characteris-tics of dense fog environment, using multi-scale weighted average Retinex algorithm of pixel gray value image adaptive en-hancement, to guide the atomization image box filter realization processing to the fog, enhanced results brightness, contrast and image structure similarity. Simulation results show that the algorithm has good defogging performance and image en-hancement performance, improves the quality of video image to mist and fog imaging rate under.