Objective Water conservancy safety is critical to national long-term stability and security. A large number of reservoirs and various dikes have been built across China. Many reservoirs were constructed in early periods under restricted technical conditions, leading to a growing number of reservoirs with potential safety hazards. Reservoir structures also deteriorate under long-term environmental impacts during service. Accurate monitoring of reservoir abutment rock mass deformation is thus essential for reservoir safety control. Conventional monitoring methods present obvious limitations. Contact measurement approaches consume massive human and material resources, rely on complex hardware systems, and fail to realize full-field monitoring. Non-contact measurement technologies, including ground-based synthetic aperture radar interferometry, global positioning system, and three-dimensional laser scanning, have respective defects that prevent large-scale engineering application.

Methods Digital image correlation technology provides multiple advantages, such as non-contact full-field measurement, simple hardware requirements, and strong adaptability. This technology computes deformation through speckle images captured before and after specimen deformation, and matches overall grayscale changes around measured points, delivering high measurement stability. It is suitable for deformation and displacement monitoring of reservoir abutment rock mass. However, digital image correlation requires multi-point calculations, resulting in low computational speed. Multi-thread parallel computing technology distributes measured points to different sub-threads for simultaneous computation based on thread quantity, which greatly improves computational efficiency. This study presents a deformation and displacement monitoring scheme for reservoir abutment rock mass based on digital image correlation and multi-thread parallel computing. To solve the problem of excessive irrelevant regions in measured images, a multi-point synchronous measurement strategy is used to ignore non-target information. This strategy is integrated with multi-thread parallel computing to accelerate the matching process of the digital image correlation method.

Results and Discussions Laboratory feasibility verification tests and field reservoir experiments are carried out. In laboratory tests, fabricated specimens are adopted to simulate deformation and displacement of reservoir abutment rock mass. In field experiments, customized targets are used to replicate the deformation behavior of reservoir abutment rock mass under in-situ conditions. Laboratory results show that the average relative error of full-field synchronous measurement is controlled within 3%. Measurement time consumption is reduced by more than 75% compared with non-multi-thread processing. In field reservoir experiments, the average relative error of full-field synchronous measurement remains within 5%, and time consumption is also reduced by more than 75%.

Conclusions Experimental results confirm that the proposed scheme provides sufficient accuracy and efficiency in both laboratory and reservoir field environments. The multi-point synchronous measurement strategy effectively eliminates interference from irrelevant image information, while multi-thread parallel computing significantly improves calculation speed. The scheme maintains stable measurement performance under complex field conditions, verifying its environmental robustness. The proposed monitoring scheme realizes reservoir abutment rock mass deformation and displacement monitoring with relatively simple hardware. It integrates high precision and high efficiency, and overcomes limitations of traditional monitoring methods. The scheme possesses strong practical application value in hydraulic engineering, and offers a reliable technical solution for long-term safety monitoring of reservoir abutment rock mass.