Research And Realization About Vectorization On Well Logging Curve

Oil and natural gas are both important energy and precious chemical resources, so it is not surprised that all the countries pay special attentions to their oil exploitation. A part of the key work of each oil field is to take full advantages of the well log information on graphs, analyze and extract the information contained, and store them properly.As the well-log information gained by down hole instrument, uphole equipment and other equipment,the borehole log mirrors the features of rock facies and depositional environment.The borehole log can forecast the ore mass distribution and the control of the spatial distribution rule of the laboratory experiments can be mainly used for such aspects as reservoir correlation,thick division,reservoir characteristic interpretation,and fine geologic description,etc.The progress of the oil exploration development can enhance the logging trace.But on the graph paper of the borehole log of the analog logging device,the amount is too much for its conservation and utilization.Therefore,it's necessary to set up the borehole log's data base to interchange the tracing patterns and the analysis of the appliance computers.The usual way to vectorizing the well log information is applying foreign software. The main tool is a mouse. How to vectorize the information rapidly and correctly remains to be an emergent unresolved problem. To solve the problem, this paper performs a thorough research on automatic vectorization. The main job this paper did is as follows:1,The research on how to extracting the back grids. This paper gives a method based on spatial recognizing that can extract back grids from normal graph effectively.2,In the research of well log curve pretreatment. The method includes repair broken curves, filtering and foreground and background segmentation of well logs.3,Tracking algorithm for curves.4,Integrate all the theories discussed in this paper, we design and realize a system that can be put into practice.Before follow-up in the curve there is a very important task is to identify the grid.Since the grid can be seen as a series of parallel and intertwined to form a straight line, so when the track will tend to misjudge the grid for the vertical extension of the well logs on the impact of vector work, so if you want fast, effective extraction of the target curve, the most direct way is to first eliminate the background grid, and then extracted on the curve. In this paper, the characteristics of the drawings in accordance with the use of log linear discriminant method to identify the grid. Linear discriminant method can effectively use the grid lines are long linear features, tilt of the drawings to avoid the impact of recognition algorithms.In the course of the grid to identify the need first to determine the elimination of the regional grid in this region to determine a binarization threshold gray partition but really do not have to be binary (for general drawing, the simple use of the area maximum and minimum gray to gray-scale average), that the gray value is less than the threshold for the black pixel points. Then the use of Hough transform to identify the region from the black point in a straight line constitutes a reasonable threshold to determine the discriminant to identify whether it is the straight-line grid, if a straight line to identify the length is less than the threshold that grid is not otherwise considered to be found in the grid and linear elimination from the drawings.Grid in order to eliminate the follow-up curve of the impact of follow-up will be identified with pure white grid to that of other pixels remain unchanged the original gray.However, logging and Grid are intertwined, will be the elimination of the grid is bound to be at the same time curve and the public part of the grid along with the elimination of lost, this will inevitably result in the fracture curve, so that the integrity of the information curve of being destroyed. In order to solve this problem, this paper proposes an approach based on the shortest distance curve matching algorithm patch. Algorithm to find the curve of the first fracture were the starting point and end point of the set, again in accordance with the principle of matching the shortest distance between start and end of the match, and then complete the matching start and end of the fracture linking curve so as to achieve the purpose of repair.In order to improve the accuracy of matching, the use of fracture orientation will start and end of the match as a supplementary information standards, to reduce or avoid errors caused by the destruction of the matching curve.Curve drawing vector tracking is the core of the process. How to log coordinates of paper quickly and automatically identify and extract the vector drawings of logging the most important, but also the largest part of the workload. Logging in order to ensure accurate and efficient tracking, this paper, the phased tracking method to complete the curve.Also in the curves based on the tracking method also studied how the vector data after the realization of the core curve of post-processing methods: refinement and peak curve compensation technology. After logging the above-mentioned processing, the amount of data compression have been fully and is conducive to well log data storage, as well as the establishment of information database for the future provides the basis for data sharing.Logging is collected by the measuring instruments to the oil wells in each of the various physical parameters change with the formation depth of the trend curve in the form of mapping the coordinates of the semi-transparent paper formed. These physical changes are random, unpredictable, and vary from wells can be expressed as a waveform curve.Waveform from the linear curve can be divided into the solid line, dotted line, points, lines, points, solid line; from the complexity of the on-line can be isolated, cross-line, such as adhesion, in which isolated line and isolated line, the dotted line isolation , isolated points and lines of points in the horizontal direction if the magnitude of said physical value, the vertical depth of said strata, these curves are "S"-shaped curve, that is, the rate and depth of the value of the one-to-one.Tracking the process of logging is to belong to the same node of a link curve, if LAG node object for the operation, the process is the follow-up search for node node.Follow-up to the choice of node and nodes below the number of degrees are closely related, if below a certain degree of node 1, its follow-up to the choice of nodes is easier; and if the node is 0 degrees or below below the number of degrees greater than 1, follow-up is not easy to choose node. In order to improve the tracking speed and tracking accuracy, in accordance with the characteristics of logging in the LAG curve established on the basis of the data structure, and use of this data structure can be a very good solution to cross-line tracking and tracking of adhesion points.Curve of the specific types of tracking methods are the basic ideas mentioned above "the curve to complete the phased tracking method", but a number of different tracking the details of the curve is slightly different.In particular paragraph in one by one from the candidate to select the part of the follow-up paragraph, we will search for a different curve to determine the choice of the region, as well as choose from the candidates follow-up paragraph to paragraph the characteristics of the choice of comparison (including the gray-scale, distance, direction, position, as well as the overlap length, segment area and the length of paragraphs, etc.). At the same time for comparison, because two lines in the same region, and there is overlap, it can not be completely separated from the individual to deal with. We are dealing with are gray curves can be compared, that is a deep curve, a simple.Even if the overlap as a result of dark basic characteristics curve was not affected, so we have adopted is to first of all one-way method by tracking dark curve, followed by the tracking method used to improve tracking of light-colored curve. Method essentially the same as one-way track, but when the scope of the specified paragraph can not find the follow-up, the need to join to find missing above procedures, the curves overlap with the dark part of the curve from the dark to find.In this paper, the former in the study mentioned above vector logging technology on the basis of a developed, entitled "Note the middle section of the monitoring data digital processing system" software. The system studied in this paper as the core of many algorithms can scan images drawing tool for processing and logging of the final disposal into a vector form.The system interfaces with the database to facilitate the use of database technology to save the results, while the system can also query the database query results and redraw the screen so that users of the results of vector analysis and application of more simple and intuitive.