针对图像复原易出现复原图像颜色视觉特征失真问题，提出基于视觉传达优化的低清晰度船舶图像复原方法。该方法通过帧扫描方法，提取低清晰度船舶图像视觉传达信息的像素二维特征；使用基于最大后验概率的高分辨率船舶图像重组方法，将所提取低清晰度船舶图像视觉传达信息的像素二维特征，进行高分辨率特征重组，并引入基于视觉颜色模型的视觉传达效果优化复原方法，优化高分辨率特征重组后复原图像的颜色视觉传达效果。实验数据验证：该方法对低清晰度海域通行船舶监控图像复原处理后，图像视觉传达效果得到明显提升，且复原后船舶图像峰值信噪比、结构相似性指数接近1，图像特征失真小；复原后图像颜色特征显著性指数最大值达1.0，颜色特征细节显著性得以改善。
A low definition ship image restoration method based on visual communication optimization is studied to address the issue of color visual feature distortion in image restoration. This method extracts pixel two-dimensional features of visual communication information from low definition ship images through frame scanning method; Using a high-resolution ship image reconstruction method based on maximum a posteriori probability, the two-dimensional features of pixels extracted from the visual communication information of low definition ship images are recombined into high-resolution features, and a visual communication optimization restoration method based on visual color models is introduced to optimize the color visual communication effect of the reconstructed image after the reconstruction of high-resolution features. Experimental data verification: After the restoration processing of low definition ship monitoring images in sea areas, the visual communication effect of the images is significantly improved, and the peak signal-to-noise ratio and structural similarity index of the restored ship images are close to 1, with small image feature distortion; The maximum value of the color feature saliency index of the restored image reaches 1.0, and the saliency of color feature details is improved.
2024,46(5): 65-68 收稿日期：2023-10-13
DOI：10.3404/j.issn.1672-7649.2024.05.012
分类号：TP391
作者简介：赵振华(1987-),女,硕士,讲师,研究方向为视觉设计
参考文献：
[1] 陈丽, 李临寒, 王世勇, 等. MMShip: 中分辨率多光谱卫星图像船舶数据集[J]. 光学精密工程, 2023, 31(13): 1962-1972.
CHEN Li, LI Lin-han, WANG Shi-yong, et al. MMShip: medium resolution multispectral satellite imagery ship dataset[J]. Optics and Precision Engineering, 2023, 31(13): 1962-1972.
[2] 曾广淼, 俞万能, 王荣杰, 等. 船舶目标重叠下马赛克图像数据增强方法研究[J]. 控制理论与应用, 2022, 39(6): 1139-1148.
ZENG Guang-miao, YU Wan-neng, WANG Rong-jie, et al. Research on mosaic image data enhancement and detection method for overlapping ship targets[J]. Control Theory & Applications, 2022, 39(6): 1139-1148.
[3] 乔志凯, 陈世才, 蒙子昕, 等. 复杂光照下的船舶水尺图像二值化新方法[J]. 中国航海, 2021, 44(4): 80-85+93.
QIAO Zhi-kai, CHEN Shi-cai, MENG Zi-xin, et al. A new method for draught line image binarization in complicated Illumination environment[J]. Navigation of China, 2021, 44(4): 80-85+93.
[4] 张恒, 杨家轩, 周洋宇, 等. 可见光视频图像中的船舶目标自适应检测[J]. 上海海事大学学报, 2021, 42(1): 33-38, 43.
ZHANG Heng, YANG Jia-xuan, ZHOU Yang-yu, et al. Adaptive detection of ship targets in visible light video images[J]. Journal of Shanghai Maritime University, 2021, 42(1): 33-38, 43.
[5] 李凯, 于洪亮, 徐轶群, 等. 基于光学遥感影像的船舶目标识别[J]. 中国航海, 2022, 45(1): 95-100.
LI Kai, YU Hong-liang, XU Yi-qun, et al. Ship target recognition from optical remote sensing image[J]. Navigation of China, 2022, 45(1): 95-100.
