Application of Image Analysis on Ground Displacement

What is it about?

The two-dimensional experiment of sandboxes for simulating ground failure has become a popular teaching tool in the current geomechanics laboratory. Most of them need cameras with high resolution and well-textured soil for digital image correlation (DIC). Those techniques do not only require a higher budget for the camera but also expensive PIV software for image analysis. Although it is more accurate and robust, it is not often fully available in every geomechanics laboratory. To provide a cheaper substitution, a simple image analyzing algorithm was developed to consistently detect the trajectory of each marker pre-embedded in a 2D sandbox. Through conducting image processing in the sequence of binary images, noise filtering and correlating the markers by smallest Euclidean distance in the sequence of images, all markers can be accurately monitored and tracked. The soil displacement of white sand under the strip footing can be consequentially measured with an acceptable accuracy. Compared to the previous DIC methods, this 2D experiment coupled with this image analyzing algorithm can measure the soil deformation for poor-textured soil. In this paper, a 2D strip footing model test is presented with newly developed methods for the measurement of ground displacement. It is a useful and straightforward experiment for investigating 2D plain-strain foundation failure.

Featured Image

Photo by Janne Simoes on Unsplash

Photo by Janne Simoes on Unsplash

Why is it important?

The two-dimensional experiment of sandboxes for simulating ground failure has become a popular teaching tool in the current geomechanics laboratory. Most of them need cameras with high resolution and well-textured soil for digital image correlation (DIC). Those techniques do not only require a higher budget for the camera but also expensive PIV software for image analysis. Although it is more accurate and robust, it is not often fully available in every geomechanics laboratory. To provide a cheaper substitution, a simple image analyzing algorithm was developed to consistently detect the trajectory of each marker pre-embedded in a 2D sandbox. Through conducting image processing in the sequence of binary images, noise filtering and correlating the markers by smallest Euclidean distance in the sequence of images, all markers can be accurately monitored and tracked. The soil displacement of white sand under the strip footing can be consequentially measured with an acceptable accuracy. Compared to the previous DIC methods, this 2D experiment coupled with this image analyzing algorithm can measure the soil deformation for poor-textured soil. In this paper, a 2D strip footing model test is presented with newly developed methods for the measurement of ground displacement. It is a useful and straightforward experiment for investigating 2D plain-strain foundation failure.

Perspectives