DSpace Repository

Fast, robust anatomical sketch recognition

Show simple item record

dc.contributor.author Le Hong Hai en_US
dc.date.accessioned 2015-01-12T10:40:33Z
dc.date.available 2015-01-12T10:40:33Z
dc.identifier.other AIT Thesis no.CS-06-14 en_US
dc.identifier.uri http://www.cs.ait.ac.th/xmlui/handle/123456789/335
dc.description Pathum Thani, Thailand : Asian Institute of Technology, 2006 en_US
dc.description 56 p. : ill. en_US
dc.description.abstract Sketches are often used in medicine. Sketchin g provides an effectiv e means of illustration. Physicians commonly use sketches when taking notes in patient records and to help convey diagnoses and treatments to patients. In medical learning, students frequently use sketches to help them think through clinical problems in individual and group problem solving. So the ability to recognize those sketches could bring benefit in many medical applications such as automated patient records and education software in medicine. With the support of tablet PC or pen-based devices, these applications become more feasible. In this thesis, I develop a new approach to recognizing anatomical sketches. The approach consists of two steps. The first step is for calculating the similarity between the sketch and a set of templates with the notice that the sketches are really different in the amount of the detail and whereas the internal details of medical sketches vary within a given catego ry, the outline is relativ ely stable. The second step uses Support Vector Machine to classify the sketch. The results of the approaches are evaluated over a number of sketches drawn by medical students. To measure the effectiveness of the approaches, the accuracies of the recognition between the system and medical students are compared.
dc.relation.ispartof Thesis no. CS-06-14 en_US
dc.relation.ispartof Asian Institute of Technology. Thesis no. CS-06-14 en_US
dc.subject Human anatomy -- Drawing -- Computer program en_US
dc.subject Robust control en_US
dc.title Fast, robust anatomical sketch recognition en_US
dc.type Thesis en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Advanced Search

Browse

My Account