This book presents a non-contact framework for measuring deformation and identifying failure mechanisms in geomaterials using digital image correlation (DIC) and computer vision. It develops a multilevel finite element-based formulation (FEM-DIC), an extended formulation (X-DIC) that resolves discontinuous displacement fields across fractures, a smoothed particle hydrodynamics estimator for full-field strain, and an automated algorithm that detects and classifies damage zones. These methods converge on a dimensionless pre-failure indicator, κ, derived from the evolution of cumulative effective strain, which identifies the onset of yielding in concrete and rock specimens before failure becomes visible. Each method is validated against numerically generated speckle images with analytically known deformation fields and then applied to uniaxial compression experiments.

Sudipta Bhattacharjee is a specialist in artificial intelligence and autonomous systems with more than 19 years of experience across computer vision, industrial AI, and safety-critical systems. He holds a Ph.D. from IIT Kharagpur, where his work on non-contact computer-vision methods for structural health monitoring and pre-failure prediction of mine roof falls forms the basis of this book. He has since held technical and leadership roles at QuEST Global, KPIT Technologies, Veoneer and Mercedes-Benz Research & Development India, and is currently a senior manager at Accenture.
Debasis Deb is a professor in the Department of Mining Engineering at IIT Kharagpur. He has held positions at Southern Illinois University Carbondale, USA, and the Korea Institute of Geoscience and Mineral Resources, South Korea. A recipient of the National Geoscience Award from the President of India, he has three books, five patents, and over 235 publications to his credit and currently serves as the president of the Indian Group of the ISRM.