This manuscript focuses on developing a workflow for determining the productivity of vineyards in a novel and innovative way, ensuring flexibility and simplicity in data acquisition, automation in the process and high-quality results, using low cost sensors. The non-invasive system proposed allows the determination of yield at cluster level by combining close-range photogrammetry and computer vision. Bunches are reconstructed in 3D from images processed with Photogrammetry Workbench software (PW) developed by the authors. Algorithms and techniques were combined to estimate the most relevant parameters in the productivity of a vineyard: volume, mass and number of berries per bunch. To validate the workflow proposed, a sample of laboratory tests based on dimensional analysis of the clusters together with the single count of berries, were analyzed to establish the groundtruth. The results achieved from the scaled models and different estimation parameters were contrasted. The results confirm the feasibility of the proposed methodology, providing scalability to a comprehensive analysis of the productivity of the vineyard and affording a constant operational improvement and proactive management.