Due to the pandemic, an increase in obesity could be observed due to the sedentary life that university students had. Overall, this is a worrying trend, particularly the rising rates of obesity in young people and their lack of physical activity. The paper aims to evaluate university students' physical activity through a computer vision prototype using Azure Kinect. This has made it possible to establish a baseline for the physical activity of young university students. For this, 50 university students from a university in Arequipa-Peru were evaluated. Two tests were carried out, one of jump height and another of bicep curl. As a result, the percentiles of the tests performed were obtained, and the prototype was developed that evaluated efficiently the correlation coefficients vertical jump test for the hip angle-Distance, it was 0.50 moderately. The Knee Angle-Distance was 0.38, considered weak; in terms of the correlation coefficients bicep curl test for the Maximum angle of flexion- Repetitions was -0.16, very weak, and for the Maximum angle of extension-Repetitions, it was -0.06, considered very weak. In conclusion, it was possible to indicate that using the computer vision prototype using Azure Kinect allows for obtaining results that specialists can use to determine the physical condition more efficiently for university students.