【摘要】 针对以往对电阻率CT图像单纯依靠直观判断的定性分析方法,提出了一种基于图像熵值的CT图像量化评价方法,从定量的角度判读图像信息的实时变化.建立了煤层底板陷落柱突水过程的简化地电模型,利用有限元方法对该地电模型的突水过程的电阻率监测进行了正演模拟,分析了电阻率CT图像熵值对突水过程的响应特征,根据图像熵值的变化信息成功捕捉到了突水前兆信息.利用图像熵值对矿井顶板突水模型试验中的动态电阻率CT图像进行评价,较突水发生时刻提前574 s捕捉到了突水前兆信息.通过综合分析数值模拟和模型试验中图像熵值响应特点,得到了"CT图像熵值的突变本质上是对导致岩层或煤层导电性根本变化的突变事件的响应"这一重要结论,发现利用CT图像熵值量化评价方法,可以在一定程度上压制初期扰动对突水前兆信息判读和捕捉的干扰,放大突变事件的影响程度,有效的提高了突水前兆信息的捕捉的成功概率和可靠性.更多还原

【Abstract】 Aiming at the qualitative analysis of resistivity CT image by intuitive judgment alone, we put forward a quantitative evaluation method based on image entropy to quantitatively analyze real-time change of image information.A simplified geo-electric model for water inrush process due to collapse columns from coal seam floor is built.And numerical forward modeling for simulating the water inrush process monitored by ERT method is executed by the Finite Element Method.The response characteristics of image entropy to the water inrush process are analyzed,by which the precursory information of water inrush is captured successfully.The image entropy method is used to evaluate the CT image in mine roof water inrush model test. And the precursory information of water inrush is captured,which is 574 s ahead of the inrush time.By analyzing the response characteristics of image entropy in numerical modeling and model test,we obtain an important conclusion,that is:"the catastrophe of image entropy is essentially the response to the catastrophe event that results in fundamental change to electric conductivity in strata or coal seam".By using the quantitative evaluation method with image entropy,the jam of the initial disturbance to analysis and capturing the precursory information is suppressed to a certain extent,and the impact of catastrophe event is amplified,making the success probability and reliability of precursory information capturing effectively improved.更多还原