Abstract: Method and System for estimating three dimensional measurements of a physical object by utilizing readings from inertial sensors is provided. The method involves capturing by a handheld unit, three dimensional aspects of the physical object. The raw recordings are received from the inertial sensors and are used to develop a raw rotation matrix. The raw rotation matrix is subjected to low pass filtering to obtain processed matrix constituted of filtered Euler angles wherein coordinates from the processed rotation matrix is used to estimate gravitational component along the three axis leading to determination of acceleration values and further calculation of measurement of each dimension of the physical object.

Abstract: The disclosure herein generally relates to the field of autonomous navigation, and, more particularly, to a diverse trajectory proposal for autonomous navigation. The embodiment discloses a hierarchical network based diverse trajectory proposal for autonomous navigation. The hierarchical 2-stage neural network architecture maps the perceived surroundings to diverse trajectories in the form of trajectory waypoints, that an autonomous navigation system can choose to navigate/traverse. The first stage of the disclosed hierarchical 2-stage Neural Network architecture is a Trajectory Proposal Network which generates a set of diverse traversable regions in an environment which can be occupied by the autonomous navigation system in the future. The second stage is a Trajectory Sampling network which predicts a fine-grained trajectory/trajectory waypoint over the diverse traversable regions proposed by Trajectory Proposal Network.

Abstract: Estimating 3D human pose from monocular images is a challenging problem due to the variety and complexity of human poses and the inherent ambiguity in recovering depth from single view. Recent deep learning based methods show promising results by using supervised learning on 3D pose annotated datasets. However, the lack of large-scale 3D annotated training data makes the 3D pose estimation difficult in-the-wild. Embodiments of the present disclosure provide a method which can effectively predict 3D human poses from only 2D pose in a weakly-supervised manner by using both ground-truth 3D pose and ground-truth 2D pose based on re-projection error minimization as a constraint to predict the 3D joint locations. The method may further utilize additional geometric constraints on reconstructed body parts to regularize the pose in 3D along with minimizing re-projection error to improvise on estimating an accurate 3D pose.

Abstract: The present disclosure provides a system and a method for stitching images using non-linear optimization and multi-constraint cost function minimization. Most of conventional homography based transformation approaches for image alignment, calculate transformations based on linear algorithms which ignore parameters such as lens distortion and unable to handle parallax for non-planar images resulting in improper image stitching with misalignments. The disclosed system and the method generates initial stitched image by estimating a global homography for each image using estimated pairwise homography matrix and feature point correspondences for each pair of images, based on a non-linear optimization. Local warping based image alignment is applied on the initial stitched image, using multi-constraint cost function minimization to mitigate aberrations caused by noises in the global homography estimation to generate the refined stitched image.

Abstract: The present disclosure provides a non-invasive, inexpensive and unobtrusive system that enables heart rate (HR) monitoring by addressing the traditionally known issues with face video based systems due to respiration, facial expressions, out-of-plane movements, camera parameters and environmental factors. These issues are alleviated by filtering, pulse modelling and HR tracking. Quality measures are defined which incorporate out-of-plane movements to define the quality of each video frame unlike existing approaches which provide a single quality for the entire video. To handle out-of-plane movement, Fourier basis function is employed to reconstruct pulse signals at affected locations. Bayesian decision theory based method performs HR tracking using previous HR and quality estimates for improved HR monitoring.

Abstract: A system and method for real time estimation of heart rate (HR) from one or more face videos acquired in non-invasive manner. The system receives face videos and obtains several blocks as ROI consisting of facial skin areas. Subsequently, the temporal fragments are extracted from the blocks and filtered to minimize the noise. In the next stage, several temporal fragments are extracted from the video. The several temporal fragments, corrupted by noise are determined using an image processing range filter and pruned for further processing. The HR of each temporal fragment, referred as local HR is estimated along with its quality. Eventually, a quality based fusion is applied to estimate a global HR corresponding to the received face videos. In addition, the disclosure herein is also applicable for frontal, profile and multiple faces and performs in real-time.

Abstract: This disclosure relates generally to a method and system for prediction of correct discrete sensor data, thus enabling continuous flow of data even when a discrete sensor fails. The activities of humans/subjects, housed in a smart environment is continuously monitored by plurality of non-intrusive discrete sensors embedded in living infrastructure. The collected discrete sensor data is usually sparse and largely unbalanced, wherein most of the discrete sensor data is ‘No’ and comparatively only a few samples of ‘Yes’, hence making prediction very challenging. The proposed prediction techniques based on introduction of temporal uncertainty is performed in several stages which includes pre-processing of received discrete sensor data, introduction of temporal uncertainty techniques followed by prediction based on neural network techniques of learning pattern using historical data.

Abstract: Feature based visual simultaneous localization and mapping (SLAM) do not produce reliable camera and structure estimates due to insufficient features in a low-textured environment. Moreover, existing visual SLAMs produce partial reconstruction when the number of 3D-2D correspondences is insufficient for incremental camera estimation using bundle adjustment. Systems and methods of the present disclosure provide edge points based monocular visual SLAM that mitigates these problems. The SLAM is initialized through a validation process. A local optimization process is provided for stable pose estimation in situations where camera tracking becomes unreliable in a very low-textured challenging environment. An efficient and reliable loop closing process that uses structural properties of edges in the frames is also provided.

Abstract: Systems and methods of the present disclosure facilitate rigid point cloud registration with characteristics including shape constraint, translation proportional to distance and spatial point-set distribution model for handling scale. The method of the present disclosure enables registration of a rigid template point cloud to a given reference point cloud. Shape-constrained gravitation, as induced by the reference point cloud, controls movement of the template point cloud such that at each iteration, the template point cloud better aligns with the reference point cloud in terms of shape. This enables alignment in difficult conditions introduced by change such as presence of outliers and/or missing parts, translation, rotation and scaling. Also, systems and methods of the present disclosure provide an automated method as against conventional methods that depended on manually adjusted parameters.

Abstract: This disclosure relates to systems and methods for solving generic inverse problems by providing a coupled representation architecture using transform learning. Convention solutions are complex, require long training and testing times, reconstruction quality also may not be suitable for all applications. Furthermore, they preclude application to real-time scenarios due to the mentioned inherent lacunae. The methods provided herein require involve very low computational complexity with a need for only three matrix-vector products, and requires very short training and testing times, which makes it applicable for real-time applications. Unlike the conventional learning architectures using inductive approaches, the CASC of the present disclosure can learn directly from the source domain and the number of features in a source domain may not be necessarily equal to the number of features in a target domain.

Abstract: Methods and systems for change detection utilizing three dimensional (3D) point-cloud processing are provided. The method includes detecting changes in the surface based on a surface fitting approach with a locally weighted Moving Least Squares (MLS) approximation. The method includes acquiring and comparing surface geometry of a reference point-cloud defining a reference surface and a template point-cloud defining a template surface at local regions or local surfaces using the surface fitting approach. The method provides effective change detection for both, rigid as well as non-rigid changes, reducing false detections due to presence of noise and is independent of factors such as texture or illumination of an object or scene being tracked for changed detection.

Abstract: Motion blur occur when acquiring images and videos with cameras fitted to the high speed motion devices, for example, drones. Distorted images intervene with the mapping of the visual points, hence the pose estimation and tracking may get corrupted. System and method for solving inverse problems using a coupled autoencoder is disclosed. In an embodiment, solving inverse problems, for example, generating a clean sample from an unknown corrupted sample is disclosed. The coupled autoencoder learns the autoencoder weights and coupling map (between source and target) simultaneously. The technique is applicable to any transfer learning problem. The embodiments of the present disclosure implements/proposes a new formulation that recasts deblurring as a transfer learning problem which is solved using the proposed coupled autoencoder.

Abstract: The present disclosure provides a non-invasive, inexpensive and unobtrusive system that enables heart rate (HR) monitoring by addressing the traditionally known issues with face video based systems due to respiration, facial expressions, out-of-plane movements, camera parameters and environmental factors. These issues are alleviated by filtering, pulse modelling and HR tracking. Quality measures are defined which incorporate out-of-plane movements to define the quality of each video frame unlike existing approaches which provide a single quality for the entire video. To handle out-of-plane movement, Fourier basis function is employed to reconstruct pulse signals at affected locations. Bayesian decision theory based method performs HR tracking using previous HR and quality estimates for improved HR monitoring.

Abstract: Feature based visual simultaneous localization and mapping (SLAM) do not produce reliable camera and structure estimates due to insufficient features in a low-textured environment. Moreover, existing visual SLAMs produce partial reconstruction when the number of 3D-2D correspondences is insufficient for incremental camera estimation using bundle adjustment. Systems and methods of the present disclosure provide edge points based monocular visual SLAM that mitigates these problems. The SLAM is initialized through a validation process. A local optimization process is provided for stable pose estimation in situations where camera tracking becomes unreliable in a very low-textured challenging environment. An efficient and reliable loop closing process that uses structural properties of edges in the frames is also provided.

Abstract: Methods and systems for change detection utilizing three dimensional (3D) point-cloud processing are provided. The method includes detecting changes in the surface based on a surface fitting approach with a locally weighted Moving Least Squares (MLS) approximation. The method includes acquiring and comparing surface geometry of a reference point-cloud defining a reference surface and a template point-cloud defining a template surface at local regions or local surfaces using the surface fitting approach. The method provides effective change detection for both, rigid as well as non-rigid changes, reducing false detections due to presence of noise and is independent of factors such as texture or illumination of an object or scene being tracked for changed detection.

Abstract: Systems and methods of the present disclosure facilitate rigid point cloud registration with characteristics including shape constraint, translation proportional to distance and spatial point-set distribution model for handling scale. The method of the present disclosure enables registration of a rigid template point cloud to a given reference point cloud. Shape-constrained gravitation, as induced by the reference point cloud, controls movement of the template point cloud such that at each iteration, the template point cloud better aligns with the reference point cloud in terms of shape. This enables alignment in difficult conditions introduced by change such as presence of outliers and/or missing parts, translation, rotation and scaling. Also, systems and methods of the present disclosure provide an automated method as against conventional methods that depended on manually adjusted parameters.

Abstract: Motion blur occur when acquiring images and videos with cameras fitted to the high speed motion devices, for example, drones. Distorted images intervene with the mapping of the visual points, hence the pose estimation and tracking may get corrupted. System and method for solving inverse problems using a coupled autoencoder is disclosed. In an embodiment, solving inverse problems, for example, generating a clean sample from an unknown corrupted sample is disclosed. The coupled autoencoder learns the autoencoder weights and coupling map (between source and target) simultaneously. The technique is applicable to any transfer learning problem. The embodiments of the present disclosure implements/proposes a new formulation that recasts deblurring as a transfer learning problem which is solved using the proposed coupled autoencoder.

Abstract: The present invention discloses a system and method for identifying a moving object in front of a vehicle by using stereo vision based technique. The invention also discloses a method and a system for determining an instantaneous distance between the moving object and the vehicle by processing the image of the moving object through a processor and by filtering and refining the image by using a Kalman filter to remove noise.

Abstract: Body joint tracking is applied in various industries and medical field. In body joint tracking, marker less devices plays an important role. However, the marker less devices are facing some challenges in providing optimal tracking due to occlusion, ambiguity, lighting conditions, dynamic objects etc. System and method of the present disclosure provides an optimized body joint tracking. Here, motion data pertaining to a first set of motion frames from a motion sensor are received. Further, the motion data are processed to obtain a plurality of 3 dimensional cylindrical models. Here, every cylindrical model among the plurality of 3 dimensional cylindrical model represents a body segment. The coefficients associated with the plurality of 3 dimensional cylindrical models are initialized to obtain a set of initialized cylindrical models. A set of dynamic coefficients associated with the initialized cylindrical models are utilized to track joint motion trajectories of a set of subsequent frames.

Abstract: The present subject matter discloses a system and a method for constructing a 3D structure. A stream of images having timestamps associated is received at first. A plurality of image sub-groups is created using the stream of images. The plurality of image sub-groups are used for extracting features of the images present in the image sub-groups. The features are used for generating point correspondences between the images. Epipolar geometries for the image sub-groups are calculated for filtering of the point correspondences, and generating filtered point correspondences. Further, 3D points representing one or more 3D point structures are generated using the filtered point correspondences. A registered 3D structure is constructed using the one or more 3D structures upon considering a scale correction and transformation between the one or more 3D structures. Further, the registered 3D structure is dense reconstructed for improving the pixel density of the 3D structure.