Performance assessment of a questionnaire input software by computer vision

Questionnaire survey is one of the important research tools for scholars to conduct quantitative information collections. For achieving fast data input, the authors have developed an automatic questionnaire input software by computer vision. This image processing system captures the questionnaire images by a high-resolution camera, and recognizes the answers inside the checkboxes in a graphical user interface swiftly. This study is designed to realize how this software performs when processing the questionnaires for data input under various conditions. In the experiments, we obtained the accuracies for the software to recognize the checkboxes filled with the check marks in commonly seen colors, as well as the accuracies to recognize the checkboxes answered with four commonly used marks. One-way ANOVA and Scheffe's test were used as the statistical analysis to find the differences between various conditions. The results can provide useful suggestions for future administration of the questionnaire survey.