To address the issues of large data volume for storage and transmission of ultrasonic guided wave full focus imaging signals for pipeline defects and low detection efficiency, a full focus imaging method based on compressive sensing and sparse matrix is studied. Firstly, six types of greedy algorithms are employed to perform compression and reconstruction on pipeline echo data. The study analyzes the influence of reconstruction algorithms on the reconstruction accuracy of simulated signals, selects the optimal reconstruction algorithm, and verifies that compressed sensing algorithms can break the constraints of the Nyquist theorem. Then, the sparsity is calculated to determine the optimal sparse basis. The measurement matrix is constructed by analyzing both the incoherence with the optimal sparse basis and the curvature effect of the pipeline. Compared with the random Gaussian matrix, the curvature-weighted measurement matrix can reduce the impact of the pipeline’s curvature effect, thereby improving the signal reconstruction accuracy and the quality of total focusing imaging. Finally, the optimal scheme is applied to conduct single-defect total focusing imaging and double-defect total focusing imaging respectively on the full-matrix data and sparse-matrix data obtained through compression and reconstruction. The results show that the sparse matrix total focusing imaging algorithm based on compressed sensing can reduce the total time for total focusing imaging and compressed sensing signal processing by 60% while ensuring accuracy. This method effectively improves imaging speed and detection efficiency, and simultaneously reduces the requirements for the hardware performance of the detection system.