ID 原文 译文
14295 针对传统的同态滤波算法增强后的图像还存在雾状模糊且颜色偏暗的问题,提出改进的同态滤波与Retinex多尺度融合的水下图像增强算法。 Aiming at the problem that theenhanced image of the traditional homomorphic filtering algorithm still has foggy blur and dark color an improved underwater image enhancement algorithm is proposed which fuses the improved homomorphic filtering with multi-scale etinex．
14296 该算法首先用基于双边滤波的单尺度Retinex算法对原始图像进行颜色校正。 Firstly the single-scale Ｒetinex algorithm based on bilateral filtering isused to correct the color of the original image．
14297 然后对修正后的图像构造相应的Butterworth陷波滤波器进行滤波。 Then the modified image is filtered by constructing acorresponding Butterworth notch filter．
14298 最后对颜色修正后的图像和同态滤波增强后的图像进行多尺度融合。 Finally the image enhanced by homomorphic filtering and the imageafter color modification are fused at multiple scales．
14299 通过实验可得,该方法能够有效地改善色偏,提高图像的清晰度。 Experimental results show that this method caneffectively suppress color offset and improve image resolution．
14300 雾霾天气下拍摄的图像,由于大气中混浊悬浮物对光的吸收和散射的影响,导致"透光"强度减弱,图像能见度严重降低,许多特征被覆盖或模糊,限制和影响了可见光视觉系统工作效用的发挥。 Images taken in foggy weather often suffer from low contrast and limited visibility due to thesubstantial presence of particles in the atmosphere which absorb and scatter light during the propagationand the efficiency of the visual system is limited and affected．
14301 研究针对低能见度天气下,可见光视觉系统工作受限的现实,从雾天图像退化模型出发,通过发掘颜色分布与透射率两者之间的约束关系构建透射率求解方程。 Aiming at the degraded performance of thevisual system in low-visibility weather the foggy image degradation model is established and an equationof transmittance is constructed by using the constraint relationship between the color distribution and thetransmittance．
14302 之后,通过对大气光值在线更新和马尔可夫随机场模型,实现对大气光及透射率的时空连续性约束,解决复杂场景以及天气条件下可见光视觉系统工作的适应性与稳健性问题。 Then spatial-temporal continuous constraint to the atmospheric light and the transmittance isrealized through online updating of atmospheric optical value and by use of Markov Ｒandom Field ( MＲF)model． The problem of adaptability and robustness of the visual system under complex environment or complexweather conditions is solved．
14303 实验结果表明,所提算法可以有效地提高低能见度条件下的成像距离,视频图像去雾后忠于原始图片颜色分布,无色彩失真,可为后续高级视觉任务提供良好的图像基础。 Experimental result shows that the proposed algorithm can effectively improve theimaging distance under low-visibility condition and the dehazed video image keeps the color distribution ofthe original image without chromatic aberration which can supply a fine image for the subsequent high-levelvision mission．
14304 针对采用含随机噪声的数据进行非线性动态系统建模无法获得准确模型参数的问题,提出了一种基于加权最小二乘支持向量机(LS-SVM)数据预处理的复合辨识方案。 Aiming at the problem that the accurate model parameters are not available for establishing anonlinear dynamic system model to the data with random noise a composite identification scheme with datapreprocessing based on weighted Least Square-Support Vector Machine ( LS-SVM) is proposed．