<p>Discontinuities, the weakest interfaces in a rock mass, greatly affect its mechanical properties and stability. Therefore, geological investigation of discontinuities is essential for characterizing the rock mass. With the development of Structure from Motion (SfM) technology, the efficiency and accuracy of 3D point cloud reconstruction for slopes have significantly improved, driving progress in intelligent geological investigation research. Although significant progress has been made in the identification of “plane” discontinuities, the existing methods for recognizing “linear” discontinuities still have bottlenecks in efficiency and accuracy in terms of fast and intelligent extraction. This study introduces a new algorithm, the Directed Path Planning (DPP) algorithm, for intelligent recognition and evaluation of linear discontinuities. The algorithm is based on the minimum spanning tree design, which can efficiently combine the color information of discontinuities and their spatially distributed density features to extract the critical path points and achieve flexible path planning. Compared to traditional manual extraction of discontinuities, the algorithm offers greater accuracy and more closely matches the actual morphology of linear discontinuities, with more detailed morphological features. Furthermore, in this study, the DPP algorithm was applied to a rocky outcrop on a slope in Alicante, southeastern Spain, where the length and spacing of the linear discontinuities were calculated using a high-precision point cloud model and compared with manually measured data. The results demonstrate that the DPP algorithm offers significant pKEAadvantages in the rapid and intelligent extraction of linear discontinuity morphology on rocky slopes, as well as in the accurate calculation and evaluation of three-dimensional features, further validating the method’s effectiveness and practicality.</p>