Zahra Sadeghi,Department of Electrical and Computer Engineering, University of Tehran, Iran
How can humans distinguish between general categories of objects? Are the subcategories of living things visually distinctive? In a number of semantic-category deficits, patients are good at making broad categorization but are unable to remember fine and specific details. It has been well accepted that general information about concepts are more robust to damages related to semantic memory. Results from patients with semantic memory disorders demonstrate the loss of ability in subcategory recognition. While bottom-up feature construction has been studied in detail, little attention has been served to top-down approach and the type of features that could account for general categorization. In this paper, I show that broad categories of animal and plant are visually distinguishable without processing textural information. To this aim I utilize shape descriptors with an additional phase of feature learning. The results are evaluated with both supervised and unsupervised learning mechanisms. The obtained results confirmed that global encoding of visual appearance of objects accounts for high discrimination between animal and plant object categories.
general categorization; visual shape descriptors; object recognition.
Shabnam Saadat1, Mark Pickering1, Diana Perriman2, Jennie M. Scarvell3 and Paul N. Smith4, 1School of Engineering and Information Technology, UNSW Canberra, Canberra, Australia,2Trauma and Orthopaedic Research Unit, The Canberra Hospital, Canberra, Australia,3Faculty of Health, University of Canberra, Canberra, Australia,4School of Medicine, The Australian National University, Canberra, Australia
Image registration has applications in different areas of medical image analysis. It can be used to assist the investigation of joint kinematics in conditions such as ligament injury, osteoarthritis, and after joint replacement. Analysing the 3D movement of joints after total knee arthroplasty surgery is crucial as the correct position and relative movement of knee implants will significantly impact the success of the surgery. However, the evaluation of the movement of the implanted components has received limited attention and most studies on this aspect are still insufficient and developing. In this paper, we propose a non-invasive and robust 3D to 2D registration method which can be used for 3D evaluations of the status of knee implants. This method addresses several challenges with regard to the registration of the implants. The experimental results show that the proposed method is not only robust but also fast.
Model to image registration, total knee arthroplasty, medical image analysis, similarity measure, Edge Position Difference.