An improved Gaussian mixture model-based adaptive point cloud registration algorithm is proposed to address the challenge of diverse and random surface appearances that hinder quick and accurate detection of deformations in coldwater pipes. The algorithm uses improved Gaussian mixture model to assign probability values for point-to-surface distances of the input point cloud, optimizes the likelihood function, and uses an adaptively adjusted parameter randomized sampling consensus algorithm to obtain the transformation matrix between the input point cloud and the design model point cloud. After feature extraction between the registered input point cloud and the design model point cloud, points exceeding a threshold are extracted separately based on normal vector features to detect deformities in the pipes. Experimental results demonstrate that this improved algorithm can rapidly and accurately detect various types of deformations in cold-water pipes with deformations less than 2mm, achieving a detection accuracy of over 99%.
Hand gesture recognition has recently grown as a powerful technical means in human-machine interaction field for control the appliances such as in home automation. However, the accuracy recognition of diverse hand gestures is still in the early stage for real-world application. In this paper, we present a new gesture recognition framework which is capable of classifying ten different hand gestures based on the input signals from surface electromyography (sEMG) sensors. The multi-channel signals of a hand motion are simultaneously captured and transmitted to a PC via Bluetooth wireless protocol. The proposed recognition framework composes of three main steps: gesture sequence segmentation, feature extraction by sparse autoencoder, and deep neural network (DNN) based classification. The advantage of the proposed approach is the automated abstract feature extraction based on sparse autoencoder method. Combined with the DNN classification technique, we could achieve a better recognition performance tested on the dataset consisting of ten types of hand gestures compared with other classification methods.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.