摘要
近几年泥石流灾害多发,而泥石流粒度分析量测任务繁重。为此,结合计算机图像处理技术,将野外实地块石照片进行计算机处理。在处理前期,将照片中块石有效地分割出来是一个最重要问题。该项技术研究在当今地质领域中是一个新兴热门领域,国内外的研究甚少,这项技术的研究现在处于萌芽阶段。有鉴于此,分析比较当今分割效果较好的3种算法并结合其他小算法,得出归一化分割算法结合分水岭算法,高低帽变换算法,beareaopen变换算法,能够较好地分割出块石,效果比较理想。
Recent years,with lots of debris flow disaster happening, the task of particle analysis and measure to debris flow is difficult in the field.Combined with computer image processing technology, the photos of rock block are going to be processed by computers for this.It is a best important question that the pohotos are segmented out correctly and effectively in the early period of processing.Now,in the geological field,this technology studied less at home and abroad is a new and hot area . The research of this technology is at the embryonic stage, and the tasks and problems that we need to face are very arduous.Analysising and comparing 3 algorithms of good segmenting effect combined with other small algorithms,the paper concludes that normalization segmentation algorithm combined with Top-Hat and Bottom-Hat and beareaopen algorithm can segment the rock block out from the pictures with good effect.
作者
秦胜伍
姚添淇
QIN Shengwu;YAO Tianqi(School of Construction Engineering,Jilin University,Changchun 130026,China)
出处
《路基工程》
2019年第1期65-70,共6页
Subgrade Engineering
