【摘要】 为应对电子扫描图像中噪声信号干扰难以捕捉电子扫描图像中形态学差异性，导致边界提取可视度不足的问题，提出考虑形态学差异的电子扫描图像目标边界提取方法。采用形态学膨胀过程、腐蚀过程处理初始电子扫描图像，去除电子扫描图像噪点。识别电子扫描图像中目标形态差异性，通过形态学开、闭运算处理电子扫描图像，提升电子扫描图像质量。采用多方向方式提取电子扫描图像目标边界，引入自适应加权对不同方向电子扫描图像目标边界提取结果融合，获得最终目标边界提取结果。实验结果表明：所提方法能够有效增强电子扫描图像质量，目标边界提取结构相似度能够稳定在98%以上，边界提取可视度均高于99%,边界完整性均高于98%,边界平滑度均高于98.5%,能够提取精确目标边界，推动电子扫描图像智能化识别进程。更多还原

【Abstract】 To cope with the interference of noise signals in electronic scanning images, it is difficult to capture morphological differences in electronic scanning images, resulting in insufficient visibility of boundary extraction.A method for extracting target boundaries in electronic scanning images considering morphological differences is proposed.Using morphological expansion and corrosion processes to process the initial electronic scan image and remove noise from the electronic scan image.Identify the morphological differences of targets in electronic scanning images, process electronic scanning images through morphological opening and closing operations, and improve the quality of electronic scanning images.Adopting a multi-directional approach to extract target boundaries from electronic scanning images, introducing adaptive weighting to fuse the extraction results of target boundaries from different directions of electronic scanning images, and obtaining the final target boundary extraction result.According to experimental analysis, the proposed method can effectively enhance the quality of electronic scanning images.The similarity of target boundary extraction structures can remain stable at over 98%,the visibility of boundary extraction is higher than 99%,the integrity of boundaries is higher than 98%,and the smoothness of boundaries is higher than 98.5%.It can extract accurate target boundaries and promote the intelligent recognition process of electronic scanning images.更多还原