Walking has become an important and sustainable mode of transportation for college students, and improving the walkability of the college campus would boost the convenience and ease of student life. This work provides a novel pedestrian accessibility approach that optimises the Walk Score method based on the frequency, diversity, and distance of students’ walking to and from public facilities, regardless of whether the students follow the COVID rules such as lane discipline, etc. Using data science approaches, we measure pedestrian density and speed. Voronoi diagrams are used to calculate pedestrian density. The Voronoi diagram allocates each point to the closest area to that point. This allows us to define a density distribution for each pedestrian and measure pedestrian speed using computer vision technology. Once the data is collected, we use Python because it has several libraries for data analysis and visualization, such as NumPy, Pandas, and Matplotlib. We visited Reva University's campus to test the tool's application, evaluate the accessibility of facility layout and walkability, and make suggestions for improvement. This assessment tool can assist urban planners and campus designers in making judgements about how to modify the facility layout of existing campuses in various regions, as well as evaluating campus plans based on the results of their walkability assessment.