我国海洋地震正在由海面探测向海底探测发展，海底地震仪OBS/GOBS的突破推动了海底地震探测的发展，提高海洋油气勘探和海洋地球结构探测的精度成为重要研究方向。与海面拖缆采集技术探测不同，海底地震采集面临高成本、高风险和精度要求等诸多考验，如何在提高数据精度的同时节约成本成为新的挑战。本申请提出基于压缩感知OBS/GOBS观测系统设计，通过采用随机采样的观测方式，在采样数远低于传统等间隔采样方案时，仍然能够达到高精度探测要求。在处理方面，采用非规则地震数据重建方案，通过改进的冷却阈值迭代算法加以GPU并行加速，提高非规则海底地震数据重构的精度和效率。最后通过模型及实际数据验证本项目方案的有效性。综上所述，本申请结合海洋地震勘探新的技术挑战和产业需求，研究基于压缩感知海底地震观测系统设计及数据重构问题，为建立高效低成本海洋地震勘探提供理论依据及技术支撑。

Marine detection is being explored from surface exploration to the seafloor. The breakthrough of Ocean Bottom Seismometor has promoted the development of seafloor seismic exploration. Improving the accuracy of offshore oil and gas exploration and deep earth structure detection has become an important research field. In terms of acquisition, unlike sea streamer seismic exploration, submarine collection faces the challenge of high cost and high risk. In order to improve the detection accuracy, it is necessary to increase the number of seismic instruments, and the cost and working hours are suddenly increased. Therefore, this project proposes a random design for OBS/GOBS observation system based on the compressed sensing. By adopting the random sampling observation method, the high precision detection requirement can still be achieved when the sampling number is much lower than the traditional equal-interval sampling scheme.In terms of processing, the irregular seismic data reconstruction scheme is adopted, by using the improved cooling threshold iterative algorithm and the GPU parallel acceleration the accuracy and efficiency of the irregular seabed seismic data reconstruction has been improved. The effectiveness of our project has been proved by both synthetic data and field data.