The purpose of this study is to estimate the volumetric distribution of discontinuities in jointed rock exposures using "Digital Image Processing (DIP)" methods. In this study, various Pre-Processing, Segmentation and Recognition methods have been applied on images of a jointed slope. Among pre-processing and segmentation algorithms, respectively, "Median Filtering" and "Niblack local binarization" method with K coefficient between -0.7 to -0.9 indicated better results. On the other hand, using "Trace Angle- Trace Quantity" histogram and engineering judgment, the value of Representative Angle (RA), Alternative Angle Limit (AAL) and Joint Set Number (JSN) can be evaluated. Similar traces are separated using mentioned parameters for spacing calculations and the rest will be classified as random joints. Applying the equation proposed by Palmstrom and using the data achieved from DIP, the Volumetric Joint Count (Jv) can be measured. Eventually, the proposed algorithms were applied on images of a jointed rock mass with known geometrical parameters. The results, obtained from DIP methods indicate proper correlation with manually surveyed parameters.
In Rock Engineering, obtaining accurate data from discontinuity traces in rock exposures play a vital role in "Engineering Judgment" and "Rock Mass Characterization". Review of different Rock Mass Characterization methods (RQD, RSR, RMR, Q, etc.) indicates that in all of them, the geometrical conditions of rock mass should be determined. Conventional data acquisition methods in are prone to different kinds of errors. Thus, during data acquisition process, new and improved methods have always been considered as a solution to emit such errors. In addition to the importance of accurate discontinuity trace map construction in rock mass characterization, precise outputs from digital imaging methods can be implemented as an input for numerical analysis softwares such as UDEC, FLAC and so on. Digital Image Processing (DIP) is one of the new techniques that despite being developed in electrical and computer engineering, but has vastly been implemented in other fields. Reid and Harrison (2000) proposed a semiautomatic method for line reconstruction whereby an operator selects a seed pixel at one of the extremities of every perceived discontinuity.[1] Kemeny et al (2003) proposed new methods using DIP and differential evolution algorithm.[2] Hadjigeorgiou et al (2003) discussed about different Line and Edge detection methods suitable for discontinuity trace map construction.[3] Moreover, Lemy and Hadjigeorgiou (2003), reviewed the general procedure to construct discontinuity map using digital imaging methods.[4] In comparison with conventional data acquisition methods in rock exposures, the new method developed by DIP has advantages as follows:
In dangerous face conditions, such as faces with unstable blocks or falling rocks, or conditions in which for any reason, lack of time for manual discontinuity mapping is inevitable, acquisition and interpretation of discontinuity network using digital images can be a great step forward.
Digital imaging algorithms can assist us in 1500 heavily jointed rock masses where distinguishing between measured and nonmeasured traces by conventional methods may be troublesome, applying DIP can avoid"Censoring" and "Truncating" biases
