Abstract

Wartegg Test is a drawing completion task designed to reflect the personal characteristics of the testers. A complete Wartegg Test has eight 4 cm x 4 cm boxes with a printed hint in each of them. The test subjects are required to use a pencil to draw eight pictures in the boxes after they saw these printed hints. In recent years, the trend of utilizing high-speed hardware and deep learning based model for object detection makes it possible to recognize hand-drawn objects from images. However, recognizing them is not an easy task, like other hand-drawn images, theWartegg Test images are abstract and diverse. Also,Wartegg Test images are multi-object images, the number of objects in one image, their distribution and size are all unpredictable. These factors make the recognition task on Wartegg Test images more difficult. In this thesis, we present a complete framework including PCC (Pearson’s Correlation Coefficient) to extract lines and curves, SLIC(Simple linear Iterative Clustering Algorithm) for the selection of key feature points, DBSCAN(Density-based spatial clustering of applications with noise) for object cluster, and finally utilize transfer learning to increase the converging speed during training and deploy YoloV3-SPP model(A deep learning network) for detecting shapes and objects. Our system produced an accuracy of 87.9% for one object detection and 75% for multi-object detection which surpass the previous results by a wide margin.