Calculating the dimensions of an object using a single camera by learning the environment
Sanjana Sanjay Mane, Chaitra Gajanan Yangandul · 2016
Dimensions like the length and breadth of an object can be useful in the field of Human Computer Interaction (HCI) for applications such as Augmented Reality. Moreover, digitally measuring the length and breadth of an object can save the tediousness of performing many day-to-day activities like geometrical calculations involving real life objects. In this paper, we talk about finding the length and breadth of an object placed in a known environment. For doing so, we first learn the environment by detecting vertical lines drawn uniformly on the ground at a distance of 5cm from each other and then we map each line to its horizontal pixel distance from the right edge of the image. With the help of Lagrange's formula for interpolation function, we can then find the dimensions of any object, subsequently placed in the same environment i.e. by keeping the camera in a fixed position. This method only requires a single camera which results in a cost effective approach. Also, the calculation complexity is reduced and the results are obtained faster with an average error rate of 2.46%.