论文部分内容阅读
随着计算机和图像处理技术的发展应用,图像识别技术已经应用于全国中药资源普查工作的各个阶段。其中,(1)在前期准备工作中,为建立统一的中药资源名录库,基于纸质文献资料采用文字识别技术,辅助进行与中药资源相关各类名录的数字化;为确定每个普查队外业调查的代表区域和样地,基于卫星遥感图像和植被图等基础数据,采用遥感图像分类等技术方法,辅助确定重点调查区域。(2)在外业调查过程中,为相对准确地获取中药材种植面积,基于遥感图像采用决策树模型、光谱特征、面向对象等方法,辅助进行中药材种植区域识别和面积估算。(3)在内业整理过程中,为相对准确的确定区域内中药资源种类,基于植物个体照片、标本照片和名称等采用图像识技术,辅助进行中药资源种类的统计汇总。(4)在成果转化应用中,基于药用资源个体和药材样品照片,开发中药资源种类识别APP和道地药材3D展示系统,辅助进行中药资源种类识别和药材鉴别特征展示。在中药资源普查工作中引入图像识别技术,辅助普查人员开展相关工作,不但能降低人工的工作量、提高工作效率,而且提高了普查成果的信息化水平和共享应用能力。随着中药资源普查工作的不断深入,图像识别技术方法在相关工作中还将发挥其特有的作用。
With the development and application of computer and image processing technology, image recognition technology has been applied to all stages of the national census of Chinese medicine resources. Among them, (1) in the preparatory work for the establishment of a unified database of traditional Chinese medicine resources, based on the use of text-based documentary literature recognition technology, assisted with the traditional Chinese medicine resources related to the various directories of digital; to determine each census team outside the industry Based on the basic data such as satellite remote sensing images and vegetation maps, the representative regions and plots of the surveyed areas were selected using techniques such as remote sensing image classification to help identify the key survey areas. (2) In order to obtain the planting area of traditional Chinese medicinal materials relatively accurately and accurately, the decision tree model, spectral characteristics and object-oriented method were used to assist the identification of area and area of Chinese herbal medicine planting. (3) In the process of housekeeping, in order to relatively accurately determine the types of traditional Chinese medicine resources in the area, an image-based technique was used to assist in the statistical summary of the types of traditional Chinese medicine resources based on photos of individual plants, specimen photos and names. (4) In the application of transformation, we developed APP and 3D display system of traditional Chinese medicine based on photos of individual and medicinal samples of medicinal resources, assisted the identification of TCM resources and the identification of medicinal materials. The introduction of image recognition technology in the census of traditional Chinese medicine resources to assist census workers to carry out related work will not only reduce the manual workload and improve work efficiency, but also improve the level of census information and the ability to share applications. With the deepening of census of traditional Chinese medicine resources, the method of image recognition technology will also play its unique role in relevant work.