In terms of the problem that ablation defects in the buffer layer of in-service high-voltage cables mostly appear at the bottom of the cable, a local source-translation computed tomography (L-STCT) method is proposed to detect defects in the water-blocking buffer layer at the bottom of in-service high-voltage cables. During the operation of L-STCT, the flat panel detector is placed close to the tested cable, estimated defect approximately exists at the maximum projection coverage angle, and projection data is collected by X-ray STCT. This paper establishes an L-STCT imaging model, analyzes scanning parameters and the maximum projection coverage angle of each data point of the scanning system, and studies their influence on image reconstruction. In addition, an experimental platform is set up to carry out a preliminary study on image reconstruction by the simultaneous iterative reconstruction technique (SIRT) algorithm. The simulation and actually experimental results show that L-STCT can detect local defects in the water-blocking buffer layer of high-voltage cables, which provides an important reference for the application of in-service high-voltage cable detection.