Separation Of Pore And Throat Based On Digital Core

This paper mainly studies separation method of pore and throat based on digital core.It mostly include two steps.Firstly,extract the medial axis of the pore space of digital core by adopting a thinning algorithm based on distance transform.Secondly,separate the pore and throat from the pore space by adopting a shape decomposition algorithm based on shape factor.There are many medial axis extraction algorithm,but most of them often produce many redundant branches influenced by boundary noise and extract the medial axis of which endpoints usually are not accurate the centrality of the pore.Even some of them are not adapted to 3D digital cores with a large amount of data because they have a large scale of computing and their complexity is relatively high.The steps of the thinning algorithm based on distance transform are as follows.First of all,calculate the distance map of the pore space of digital core,and find the apparent pore centrality which have the local minimum boundary distance value.Next,thin the pore space of digital core by adopting the thinning algorithm,and monitor the thinning process,so that two endpoints are both apparent pore centrality and maintain the connectivity of the pore space,until the pore space is transformed to curves,we called medial axis,which are 1 pixel width.At last,cut off the redundant branches generated when extracting the medial axis.Adopting the thinning algorithm based on distance transform extracts the medial axis of which endpoints are accurate,and is little influenced by boundary noise,and can cut off the redundant branches easily and effectively.Besides,it has a small amount of computation and relatively low complexity,and it is really adapted to 3D digital cores with a large amount of data very much.Now there are also many shape decomposition algorithm.Some of them have a good effect,and accord with human vision.But most of them can not predict how many parts the object will decomposed into,and can not pay more attention to outstand the shape of pore than throat because people concern the shape of pore more than throat according to the habit.Besides,most of them are only adapted to 2D image,they have a large scale of computing and their complexity is relatively high.The steps of the shape decomposition algorithm based on shape factor are as follows.Firstly,establish the segment plane,every medial axis through every pore centrality is corresponding to a segment plane.Secondly,calculate the shape factor of the separated pore.Finally,obtain the segment plane combination which make the shape factor of the separated pore reach the maximum value by changing locations of segment planes.Decomposition algorithm based on the shape factor has a great effect in accordance with human vision,and has a standard rule for assessing the shape decomposition quality.Moreover,our method is prior to decompose pore,is apt to outstand pore shapes.And our method is predictable,the numbers of decomposed pore and throat are exactly the number of pore centrality and medial axis.In addition,Decomposition algorithm based on the shape factor has a small amount of computation,and it is pretty applicable to 3D digital cores.