3 resultados para Pedestrian Overpasses.

em Indian Institute of Science - Bangalore - Índia


Relevância:

20.00% 20.00%

Publicador:

Resumo:

Computer Vision has seen a resurgence in the parts-based representation for objects over the past few years. The parts are usually annotated beforehand for training. We present an annotation free parts-based representation for the pedestrian using Non-Negative Matrix Factorization (NMF). We show that NMF is able to capture the wide range of pose and clothing of the pedestrians. We use a modified form of NMF i.e. NMF with sparsity constraints on the factored matrices. We also make use of Riemannian distance metric for similarity measurements in NMF space as the basis vectors generated by NMF aren't orthogonal. We show that for 1% drop in accuracy as compared to the Histogram of Oriented Gradients (HOG) representation we can achieve robustness to partial occlusion.

Relevância:

20.00% 20.00%

Publicador:

Resumo:

Shoe-mounted inertial sensors offer a convenient way to track pedestrians in situations where other localization systems fail. This tutorial outlines a simple yet effective approach for implementing a reasonably accurate tracker. This Web extra presents the Matlab implementation and a few sample recordings for implementing the pedestrian inertial tracking system using an error-state Kalman filter for zero-velocity updates (ZUPTs) and orientation estimation.

Relevância:

10.00% 10.00%

Publicador:

Resumo:

Comparison of reflectivity data of radars onboard CloudSat and TRMM is performed using coincident overpasses. The contoured frequency by altitude diagrams (CFADs) are constructed for two cases: (a) only include collocated vertical profiles that are most likely to be raining and (b) include all collocated profiles along with cloudy pixels falling within a distance of about 50 km from the centre point of coincidence. Our analysis shows that for both cases, CloudSat underestimates the radar reflectivity by about 10 dBZ compared to that of TRMM radar below 15 km altitude. The difference is well outside the uncertainty value of similar to 2 dBZ of each radar. Further, CloudSat reflectivity shows a decreasing trend while that of TRMM radar an increasing trend below 4 km height. Basically W-band radar that CloudSat flies suffers strong attenuation in precipitating clouds and its reflectivity value rarely exceeds 20 dBZ though its technical specification indicates the upper measurement limit to be 40 dBZ. TRMM radar, on the other hand, cannot measure values below 17 dBZ. In fact combining data from these two radars seems to give a better overall spatial structure of convective clouds.