Abstract—This paper proposes a novel touchless visualization system for computer aided surgery (CAS), which can control and manipulate the patient’s 3D anatomy model without contact through the use of Kinect-based gesture recognition technology. Real-time visualization is important in surgery, particularly during the operation. But traditional input devices are reliant on physical contact, which are ill-suited for non-sterile conditions. The depth and skeleton information from Kinect are effectively utilized to produce markerless hand extraction. Based on this representation, histogram of oriented gradients (HOG) features and principal component analysis (PCA), are used to recognize hand gestures. We developed a new system, which can visualize 3D medical image with L form screen and 9 kinds of simple touchless single-handed interactions. Experiments show that the proposed system is able to achieve high accuracy.

Index Terms—Computer aided surgery(CAS), kinect, touchless visualization, medical image.

Jiaqing Liu, Ryoma Fujii, Yutaro Iwamoto, and Yenwei Chen are with Graduate School of Information Science and Engineering, Ritsumeikan University, Shiga, Japan.
Tomoko Tateyama is with Department of Computer Science, Hiroshima Institute of Technology, Hiroshima Japan.

Cite:Jiaqing Liu, Ryoma Fujii, Tomoko Tateyama, Yutaro Iwamoto, and Yenwei Chen, "Kinect-Based Gesture Recognition for Touchless Visualization of Medical Images," International Journal of Computer and Electrical Engineering vol. 9, no. 2, pp. 421-429, 2017.