The Technical Study Of Automatic Recognition Of Geological Fracture From Borehole Images

The research of automatic recognition of geological fracture and aperture is an important content in borehole image interpretation, and in research of reservoir describe and reservoir's Heterogeneous. Fracture provides not only accumulation space for oil but also channels for migration of hydrocarbons. It is a key factor to control the production of fracture reservoir. Study of automatic recognition of geological fracture from borehole images will be beneficial to build and develop the subject of borehole images interpretation and have important application value in site.This paper focuses on the theoretical analysis of image processing in research field of fracture. We firstly study the characterization of fracture in borehole images, and then scan it into computer, save it in Bitmap-file formats. There is a lot of inevitable random noise and insufficient exposure in images. In order to eliminate noise and enhance contrast of image, we apply linear smoothing filter and gray-stretch methodology in the procedure of image pre-processing before image analysis. After this, the information of fracture is effectually exposed from background.In this paper, we optimally choose Sobel operator and Kirsch operator in edge detection of fracture. An alternative and largely innovative approach is offered, employing a consistent and automatic reduction of voluminous borehole image data to pertinent image fracture's edge attributes, which is base on Histogram-based segmentation techniques. Otherwise principle and algorithm, such as Hough transform and mathematical morphology and edge detection and following are detailedly discussed in this paper.Edge detection and following algorithm and Hough transform are particularly applied in automatic recognition of vertical fracture and incline fracture. Edge detection and following algorithm are applied in automatic identification the presence of a lot of apertures in carbonates. The experiment examples indicate that these algorithms are easy in use and insensitive for noise and occlusive in complex fracture aperture detection problems.Because of the difficulty of calculating the fracture aperture parameters and complication of numerical simulation and dependence on image tools in well logging, mathematical statistics which base on image processing is employed in order to quantitatively calculate fracture parameter, such as apparent width and apparent long and apparent Area and porosity-area ratio, and aperture parameters, such as mean of opening and porosity-area ratio. A new effectual technique based on image processing theory by which parameters of fracture are calculated is developed. The experiment research examples indicate that this algorithm is effectual and original.