Computer-based reconstruction models can be used to approximate urban environments. These models are usually based on several mathematical approximations and the usage of different sensors, which implies dependency on many variables. The sensitivity analysis presented in this paper is used to weigh the relative importance of each uncertainty contributor into the calibration of a panoramic camera–LiDAR system. Both sensors are used for three-dimensional urban reconstruction. Simulated and experimental tests were conducted. For the simulated tests we analyze and compare the calibration parameters using the Monte Carlo and Latin hypercube sampling techniques. Sensitivity analysis for each variable involved into the calibration was computed by the Sobol method, which is based on the analysis of the variance breakdown, and the Fourier amplitude sensitivity test method, which is based on Fourier’s analysis. Sensitivity analysis is an essential tool in simulation modeling and for performing error propagation assessments.
This paper describes a computer vision system designed to perform an inventory of traffic signals. The system consists
of five Ethernet synchronized cameras; the acquisition strategy allows us to take one image per camera every other
second. We then use those five images to generate a panoramic image each second. Signal detection and recognition is
carried out offline. Detection of traffic signal is done in the panoramic image using the Hough transform and
enhancement of HSV color space. Traffic signal recognition is made by a combination of Haar wavelet and violates Jones
classifier. Finally, we present experimental results using a database of one hundred images.
In this work we present results of measuring deformations on cylindrical objects. The proposed technique is
based on panoramic vision principles with convex mirrors, particularly paraboloid mirrors. ESPI techniques are
then combined with this panoramic vision system to create a technique suitable of being applied on cylindrical
objects. Some results obtained with first-approach mirrors are shown. The system is composed of two modules,
illumination and capture, each one needs a paraboloid mirror. Nevertheless this is not the only possible setup
when using convex mirrors, an alternative setup is also proposed but is not studied experimentally. Results
show the feasibility of the system to determine full-field deformations inside a cylindrical object. The systems is
currently being patented and points to an attractive solution in cylindrical or panoramic geometries.