图像信息熵约束的浅地层层界划分方法

Demarcation of the sub-bottom layers based on image information entropy constraint

摘要:

为实现快速、精确、自动化、智能化的海底浅地层层界提取, 克服传统浅地层层界在复杂海洋环境下提取时的低效、模糊、主观性等缺点, 提出一种基于图像信息熵约束的浅地层层界划分方法.首先, 将浅剖图像分割为不同区块；然后, 在不同区块计算信息熵, 并结合钻孔数据, 建立信息熵与显著性参数关系模型；最后, 据此模型对整个浅剖图像进行层界划分.研究表明, 该方法克服了现有方法的不足, 实现了浅地层剖面层界的自适应、准确划分, 试验中取得了与钻孔层界深度、厚度同量级的精度.由此可知采用图像信息熵约束进行层界提取, 可以实现浅地层层界提取的自动化与智能化

Abstract:

To address the issue of the sub-bottom profile layer extraction in complex circumstance, this paper proposes a new demarcating method based on constraint of image information entropy. Firstly, the image of sub-bottom is divided into different blocks; then, the information entropy in each block is calculated and a relation model of information entropy and significant parameters are established according to drilling data; finally, the whole sub-bottom profiling is demarcated according to the model. It is revealed that this method has overcome the shortcomings of existing methods, realized the self-adapting and exacted demarcation of sub-bottom layers. The experiment has gained the same accuracy as the depth and thickness of layers got by drilling data.