复杂背景下的树木图像提取

Study surveys on tree image extraction in a complex background

树木图像提取是将树木地面摄影图像中的单株树木与其周围景物分离的技术,为计算机技术在林业中的应用提供基础数据和技术支撑。在自然场景中拍摄的背景具有不确定性的树木图像,其提取是一项探索性很强的工作,具有重要的实用价值和现实意义。本文分析了复杂背景下树木图像的特点和提取的难点等问题;把近年来出现的树木图像提取方法按照图像分割和自然图像抠图分成两大类,并分别对每种方法中的典型算法进行了详细阐述。在此基础上,对图像分割与自然图像抠图技术在树木图像提取应用中的优势和局限性进行了分析和比较,提出把图像分割和自然图像抠图技术合理结合,针对树木图像特征,设计出准确、快速、实时的树木图像提取技术是未来研究的重点和难点。

Tree image extraction is an approach to separate an object tree from its surrounding landscape in a photograph,for the purpose of providing basic data and technical support for the application of computer technology in forestry.The uncertainty and complexity of the background in a photograph shot in a natural environment make the extraction of tree images challenging.We analyzed the issues and difficulties of tree image extraction in a complex background.The extraction methods are divided into two groups,i.e.image segmentation and natural image matting.We expound on these two methods and discuss their advantages and limitations.A combination of these two methods could become a suitable approach to the extraction of tree images.The development of an accurate,rapid and real-time tree image extraction technology is the focus in future research.