<p>Optically Induced Dielectrophoresis (ODEP) has been widely used in biomedical applications such as cell sorting and cell capture because of its operational flexibility and low cellular damage. However, existing automated ODEP methods often lack effective control of non-target cells, which may reduce manipulation performance in multicellular environments. To address this problem, this study proposes an automated cell manipulation method integrating ODEP, image processing, static optical traps and the A-star algorithm. Cells are first identified and localized from microscopic images. Non-target cells are then constrained by static optical traps and treated as static obstacles during path planning. Based on the detected cell positions, obstacle avoiding paths are generated and converted into executable optical patterns. Experiments were performed using yeast cells under a frequency of 1 kHz, a voltage of 2 V, and a light spot velocity of 5 μm/s. The results showed that the target cells followed the planned obstacle-avoiding paths and reached the designated destinations, while the non-target cells remained confined within their corresponding static optical-trap regions. The success rates of repeated single-cell directed transport and two-cell convergence experiments were approximately 90% and 80%, respectively. Non-target cells showed mean displacements of 0.68 μm (n = 30, single-cell) and 3.74 μm (n = 14, two-cell). Although larger in the two-cell experiment, none escaped optical traps or interfered with target manipulation. This work demonstrates the feasibility of combining static optical confinement with automated path planning for cell manipulation in multicellular fields of view and provides a basis for further studies involving denser and more complex cellular environments.</p>