This paper aims to provide a comprehensive survey of recent advancements in modelling and autonomous manipulation non-rigid objects. It first summarizes the advances sensing such objects with focus on describing methods technologies used measure their shape estimate material physical properties. Formal representations considered predict deformation resulting from are then investigated. The third part provides planning control strategies exploited operate dexterous robotic systems while...

The distributed measured data in large regions and remote locations, along with a need to estimate climatic for point sites where no have been recorded, has encouraged the implementation of spatial interpolation techniques. Recently, increasing use artificial intelligence become promising alternative conventional deterministic algorithms interpolation. present study aims evaluate some machine learning-based against strategies interpolating soil temperature from region southeast Canada an...

This paper discusses the design and implementation of a framework that automatically extracts monitors shape deformations soft objects from video sequence maps them with force measurements goal providing necessary information to controller robotic hand ensure safe model-based deformable object manipulation. Measurements corresponding interaction at level fingertips position three-finger are associated contours deformed tracked in series images using neural-network approaches. The resulting...

Soil temperature is a critical parameter in soil science, agriculture, meteorology, hydrology, and water resources engineering, its accurate cost-effective determination prediction are very important. Machine learning models widely employed for surface, near-surface, subsurface predictions. The present study properly designed one-dimensional convolutional neural network model to predict the hourly at depth of 0–7 cm. annual input dataset this included eight climatic features. performance was...

Existing skeleton-based human action classification models rely on well-trimmed action-specific skeleton videos for both training and testing, precluding their scalability to real-world applications where untrimmed exhibiting concatenated actions are predominant. To overcome this limitation, recently introduced segmentation involve un-trimmed into end-to-end training. The model is optimized provide frame-wise predictions any length of testing videos, simultaneously realizing localization...

Controlling robotic interventions on small devices creates important challenges the sensing stage as resolution limitations of noncontact sensors are rapidly reached. The integration haptic to refine information provided by vision appears a very promising approach in development autonomous systems because it reproduces multiplicity sources used humans. This paper discusses an intelligent multimodal sensor system developed enhance control manipulations three-dimensional (3-D) objects....

This paper presents an approach for calibrating a network of Kinect devices used to guide robotic arms with rapidly acquired 3D models. The method takes advantage the rapid measurement technology embedded in sensor and provides registration accuracy within range depth measurements provided by this technology. internal calibration between color is also presented. resulting system developed inspect large objects, such as vehicles, positioned enlarged field view created RGB-D sensors.

Rheumatoid arthritis (RA) is an inflammatory disease causing pain, swelling, stiffness, and loss of function in joints; it difficult to diagnose early stages. An diagnosis treatment can delay the onset severe disability. Infrared (IR) imaging offers a potential approach detect changes degree inflammation. In 18 normal subjects 13 patients diagnosed with Arthritis (RA), thermal images were collected from joints hands, wrists, palms, knees. Regions interest (ROIs) manually selected all parts...

Probabilistic occupancy grids have proved to be very useful for workspace modeling in 2D environments. Due the expansion of computational load, this approach was not tractable mapping a 3D environment real applications. In paper, original grid scheme is revisited and generic closed-form function introduced avoid numerical computation probabilities range sensor with Gaussian error distribution. Occupancy are computed stored multiresolution octree improved performance compactness. models built...

For the first phase of a large project, we used an infrared imaging camera (thermograph) to obtain accurate measurements body temperature in joints twelve human normal subjects (control group) and for thirteen patients who had been diagnosed with rheumatoid arthritis (RA) by rheumatologist. The ultimate goal is create low cost effective method diagnose early synovitis. Temperature hands were analyzed order statistics. Results show significant differences between control every joint hand...

As an extension to classical structured lighting techniques, the use of bi-dimensional pseudo-random color codes is explored perform range sensing with variable density from a stereo calibrated rig and projector. Pseudo-random are used create artificial textures on scene which extracted grouped in confidence map ensure reliable feature matching between pairs images taken two cameras. Depth estimation performed corresponding points progressive refinement as pattern projection marched over...

This paper presents a method for the visual detection of parts interest on outer surface vehicles. The proposed combines computer vision techniques and machine learning algorithms to process images lateral views automobiles. aim this approach is determine location set car in ordinary scenes. can be used intelligent transportation industry construct advanced monitoring security applications. key contributions work are introduction methodology locate multiple patterns cluttered scenes vehicles...

This paper introduces a new hybrid control architecture for solving the navigation problem of mobile robot in an unknown dynamic environment based on actual-virtual target switching strategy. is combination deliberative and reactive architectures which consists three layers: modeling, planning reaction. The produces collision-free with shortest-distance path, while using generates safe time minimal path. proposed approach differs from previous ones its integration architecture, techniques...

This paper discusses the importance of providing non- invasive techniques in analysis complex movements performed by musicians and athletes. Current gesture systems are insufficient do not succeed quality performance measurements without imposing strict environmental operational constraints on individuals. Computer vision offers means with which such can be made possible impeding or compromising integrity measurements. The following study compares some modern image segmentation their...

Abstract The continuous rise in the number of vehicles circulation brings an increasing need for automatically and efficiently recognizing vehicle categories multiple applications such as optimizing available parking spaces, balancing ferry loads, planning infrastructure managing traffic, or servicing vehicles. This paper explores use human visual attention mechanisms to identify a set features that allows fast automated classification based on images taken from 6 viewpoints. Salient...

The paper presents design and implementation issues of a natural gesture user interface for real-time hand recognition in depth data collected by Kinect sensor. hands are first segmented based on novel algorithm improves the scanning time order to identify pixel contour within this space. Starting from pixel, directional search allows identification entire k-curvature is then employed locate fingertips over it. Dynamic warping (DTW) guides selection candidates also recognizes gestures...

Necrotizing enterocolitis (NEC) is a severe condition in neonates, typically involving inflammation the small intestine. In this paper, novel automated image acquisition and analysis system combining an infrared thermal camera with RGB-D sensor proposed for detection of NEC preterm newborns. Calibration procedures are defined to ensure frame synchronization observation consistency among color, depth images. Segmentation extraction body automatically performed on help additional color...