4:30 PM - 4:45 PM
[MIS16-05] Demonstration of a prototype automatic feature recognition algorithm of strata boundaries in the Izu-Oshima Chisou Daisetsudanmen outcrop
Keywords:Image recognition, Automation, Enhancement of geological survey method, Autonomous probe
The purpose of this research is to establish the in-situ geological survey method by replacing the traditional outcrop description and classification process, which is strongly relying on the experts’ knowledge and experience, with computational processing using an automatic recognition algorithm for lithologic features. The first phase of this research is to improve the efficiency of geological surveys on the Earth and to increase the possibility of making new discoveries. In the second phase, key technologies will be developed for automatic selection of sites for detailed scientific data acquisition in space exploration of underground cavities on the Moon and Mars, and for data compression and high-efficiency transfer using edge computing, in order to realize an efficient autonomous exploration system.
As an initial demonstration, prototype code of an automatic recognition algorithm was demonstrated in a geological survey of the outcrop of the airfall tephra beds exposing on the “Chisou Daisetsudanmen” outcrop in Izu-Oshima as an example of the simple stratified beds without major folding and unconformity.
(1) Automated recognition algorithm of stratum boundary lines is developed, the algorithm consists of the following functions: image preprocessing to extract only the main formation structure, recognition of boundaries with large image intensity gradients, and removal of unnecessary boundary lines.
(2) Automated scientific feature recognition web-service system for strata boundaries recognition is developed to realize the execution of feature recognition from any device such as tablet computers.
(3) Developed system has been successfully demonstrated for the practical geological survey. It can provide the highlighted image of automatically recognized strata boundaries for the manual description of boundaries by the survey researchers. Strata boundary detection image has successfully used as input data for automated geological columnar section generation tool developed by collaborating researcher.
As a result of the demonstration, the main strata boundaries are shown to be recognized appropriately by the automatic feature recognition algorithm of strata boundaries. It is expected that the developed code can be applied for various objects on Izu-Oshima Chisou Daisetsudanmen outcrop. The application of the developed generalized image recognition algorithm based on the quantitative image processing is promising approach to realize the geological survey methods that are less susceptible to inexperience and variations in the quality of the surveyor’s work.
In the future, the robustness of the feature recognition algorithm for various types of outcrops will be improved, and the function of the extraction of the strata boundary lines as a series of line segment and the grain size recognition algorithms will be investigated. In addition to boundary line recognition, the texture properties for each image segments will be utilized for the further complicated feature recognition for the automated generation of the geological columnar section and other objectives to enhance geological survey method.
As an initial demonstration, prototype code of an automatic recognition algorithm was demonstrated in a geological survey of the outcrop of the airfall tephra beds exposing on the “Chisou Daisetsudanmen” outcrop in Izu-Oshima as an example of the simple stratified beds without major folding and unconformity.
(1) Automated recognition algorithm of stratum boundary lines is developed, the algorithm consists of the following functions: image preprocessing to extract only the main formation structure, recognition of boundaries with large image intensity gradients, and removal of unnecessary boundary lines.
(2) Automated scientific feature recognition web-service system for strata boundaries recognition is developed to realize the execution of feature recognition from any device such as tablet computers.
(3) Developed system has been successfully demonstrated for the practical geological survey. It can provide the highlighted image of automatically recognized strata boundaries for the manual description of boundaries by the survey researchers. Strata boundary detection image has successfully used as input data for automated geological columnar section generation tool developed by collaborating researcher.
As a result of the demonstration, the main strata boundaries are shown to be recognized appropriately by the automatic feature recognition algorithm of strata boundaries. It is expected that the developed code can be applied for various objects on Izu-Oshima Chisou Daisetsudanmen outcrop. The application of the developed generalized image recognition algorithm based on the quantitative image processing is promising approach to realize the geological survey methods that are less susceptible to inexperience and variations in the quality of the surveyor’s work.
In the future, the robustness of the feature recognition algorithm for various types of outcrops will be improved, and the function of the extraction of the strata boundary lines as a series of line segment and the grain size recognition algorithms will be investigated. In addition to boundary line recognition, the texture properties for each image segments will be utilized for the further complicated feature recognition for the automated generation of the geological columnar section and other objectives to enhance geological survey method.