Determining the mechanical properties of structures can be quite a complex and time-consuming process, requiring extensive preparation of the test subject for testing. Due to significant advancements in artificial intelligence, computer vision, and machine learning, this process can be simplified and accelerated. This thesis provides a detailed presentation and explanation of the procedure for establishing and training a convolutional neural network model that, based on input parameters, returns the mechanical properties of the test subject. Using existing Python packages for computer vision, we obtained the necessary input parameters to acquire the desired mechanical properties.
|
