支持场景深度图像重建的知识表达及应用

Knowledge representation and application of objects supporting scene range image reconstruction

随着知识工程的发展,知识已经广泛应用于船舶的设计、建造、修理等过程中.为了改善和提高逆向建模的效率和质量,提出基于知识的逆向建模处理方法,即将知识工程引入逆向重建过程中.以液舱深度图像数据为研究对象,研究液舱的相关知识表达及其空间关系表达,并采用基于XML的知识表示方法,将知识应用在液舱深度数据的逆向建模过程中.以特征提取为例,在知识的引导作用下,在对比精度为10-4mm的前提下,提取精度达到90%以上,同时也使得噪声数据、缺失数据的影响得到有效的处理.实验证明:所提供的处理方法不仅可以使建模质量得到保障,并且可使建模效率得到大幅度的提高.

With the development of knowledge engineering, knowledge has been widely used in design, manufacturing and overhaul of ships. To improve the efficiency and quality of reverse modeling, processing method of reverse engineering based on knowledge is proposed, which means bringing knowledge engineering into reverse modeling process. This paper studied the representation of the related knowledge and the spatial relationship of liquid tanks based on knowledge representation of XML and applied for the reverse modeling of liquid tanks. Take feature extraction for example,on the premises of comparative accuracy of 10^-4 mm, the extraction preci- sion has increased 90 % and the influence of noise data and deficiency data have been well-controlled. The experiment shows that the approach provided can not only ensure the quality of modeling, but also enhance modeling efficiency greatly.