There are several shortcomings in the assessment of human-induced vibration in walkways, including a focus on structural response rather than pedestrian comfort, the reliance on structural vibration response to evaluate pedestrian comfort, and the limitations of data collection methods. To address these problems, this paper proposes a more comprehensive approach, namely human-induced vibration serviceability assessment of full-path footbridge based on computer vision with source-path-receiver. The proposed method captures video sequences of both the footbridge and pedestrian movements under pedestrian excitation using computer vision techniques, and then utilizes the segmental optical flow method and the MMTracking algorithm to obtain the vibration response of both. The acceleration responses of the pedestrians obtained from the above extractions and transformations are used as an evaluation index for the vibration comfort of the pedestrian bridge in terms of the root mean square value of acceleration. To validate the feasibility and accuracy of this method, experiments were conducted on the pedestrian bridge in the laboratory. The results show that the computer vision technology can accurately and contactlessly capture the pedestrian dynamic information of the footbridge, which is more reasonable than the conventional method which only evaluates the vibration comfort of the footbridge based on the structural vibration response. By addressing the shortcomings of current assessment standards and methods, this approach provides a more comprehensive and accurate means of evaluating the vibration serviceability of footbridges, considering both the structural response and the actual experience of pedestrians.