Patricio A. Vela

A comparative study of efficient initialization methods for the k-means clustering algorithm

K-means is undoubtedly the most widely used partitional clustering algorithm. Unfortunately, due to its gradient descent nature, this algorithm is highly sensitive to the initial placement of the cluster centers. Numerous initialization methods have been proposed to address this problem. In this paper, we first present an overview of these methods with an emphasis on their computational efficiency. We then compare eight commonly used linear time complexity initialization methods on a large and diverse collection of data sets using various performance criteria. Finally, we analyze the experimental results using non-parametric statistical tests and provide recommendations for practitioners. We demonstrate that popular initialization methods often perform poorly and that there are in fact strong alternatives to these methods.

Personnel tracking on construction sites using video cameras

This paper discusses the possibility of- and need for-tracking workforce on construction jobsites using video cameras. An evaluation of algorithms and their associated results is presented. The principal objective of this paper is to test and demonstrate the feasibility of tracking workers from statically placed and dynamically moving cameras. This paper also reviews existing techniques to monitor workforce and describes areas where this work might be useful in engineering applications. The main difficulties associated with tracking on a construction site is the significant amount of visual clutter, the changing photometric visual content throughout the course of a day, and the presence of occluding and moving obstacles. The tracking of workers within the field of view of the camera will involve four tracking techniques, density mean-shift, Bayesian segmentation, active contours, and graph-cuts. Typical construction site video will be processed using the proposed algorithms and analyzed to determine the most appropriate tracking method for the video presented.

Tracking multiple workers on construction sites using video cameras

This paper proposes a tracking scheme for tracking multiple workers on construction sites using video cameras. Prior work has compared several contemporary tracking algorithms on construction sites and identified several difficulties, one of which included the existence of interacting workforce. In order to address the challenge of multiple workers within the cameras field of view, the authors have developed a tracking algorithm based upon machine learning methods. The algorithm requires several sample templates of the tracking target and learns a general model that can be applied to other targets with similar geometry. A parameterized feature bank is proposed to handle the case of variable appearance content. The tracking initialization has been discussed for different types of video cameras. A multiple tracking management module is applied to optimize the system. The principal objective of this paper is to test and demonstrate the feasibility of tracking multiple workers from statically placed and dynamically moving cameras.

Vision-Based Tower Crane Tracking for Understanding Construction Activity

AbstractVisual monitoring of construction worksites through the installation of surveillance cameras has become prevalent in the construction industry. These cameras also are useful for automatic observation of construction events and activities. This paper demonstrates the use of a surveillance camera for assessing tower crane activities during the course of a workday. In particular, it seeks to demonstrate that the crane jib trajectory, together with known information regarding the site plans, provides sufficient information to infer the activity states of the crane. The jib angle trajectory is tracked by using two-dimensional to three-dimensional rigid pose tracking algorithms. The site plan information includes a process model for the activities and site layout information. A probabilistic graph model for crane activity is designed to process the track signals and recognize crane activity as belonging to one of two categories: concrete pouring and nonconcrete material movement. The experimental result...

Construction performance monitoring via still images, time-lapse photos, and video streams

Timely and accurate monitoring of onsite construction operations can bring an immediate awareness on project specific issues. It provides practitioners with the information they need to easily and quickly make project control decisions. Despite their importance, the current practices are still time-consuming, costly, and prone to errors. To facilitate the process of collecting and analyzing performance data, researchers have focused on devising methods that can semi-automatically or automatically assess ongoing operations both at project level and operation level. A major line of work has particularly focused on developing computer vision techniques that can leverage still images, time-lapse photos and video streams for documenting the work in progress. To this end, this paper extensively reviews these state-of-the-art vision-based construction performance monitoring methods. Based on the level of information perceived and the types of output, these methods are mainly divided into two categories (namely project level: visual monitoring of civil infrastructure or building elements vs. operation level: visual monitoring of construction equipment and workers). The underlying formulations and assumptions used in these methods are discussed in detail. Finally the gaps in knowledge that need to be addressed in future research are identified.

Automatic tracking of flying vehicles using geodesic snakes and Kalman filtering

This paper describes a tracking algorithm relying on active contours for target extraction and an extended Kalman filter for relative pose estimation. This work represents the first step towards treating the general problem for the control of several unmanned autonomous vehicles flying in formation using only local visual information. In particular, we only allow on-board passive sensing. The problem is an excellent paradigm for studying the use of visual information in a feedback loop, the central theme of controlled active vision.

Reproducing Kernel Hilbert Space Approach for the Online Update of Radial Bases in Neuro-Adaptive Control

Classical gradient based adaptive laws in model reference adaptive control for uncertain nonlinear dynamical systems with a Radial Basis Function (RBF) neural networks adaptive element do not guarantee that the network weights stay bounded in a compact neighborhood of the ideal weights without Persistently Exciting (PE) system signals or a-priori known bounds on ideal weights. Recent work has shown, however, that an adaptive controller using specifically recorded data concurrently with instantaneous data can guarantee such boundedness without requiring PE signals. However, in this work, the assumption has been that the RBF network centers are fixed, which requires some domain knowledge of the uncertainty. We employ a Reproducing Kernel Hilbert Space theory motivated online algorithm for updating the RBF centers to remove this assumption. Along with showing the boundedness of the resulting neuro-adaptive controller, a connection is also made between PE signals and kernel methods. Simulation results show improved performance.

Optimized selection of key frames for monocular videogrammetric surveying of civil infrastructure

Videogrammetry is an inexpensive and easy-to-use technology for spatial 3D scene recovery. When applied to large scale civil infrastructure scenes, only a small percentage of the collected video frames are required to achieve robust results. However, choosing the right frames requires careful consideration. Videotaping a built infrastructure scene results in large video files filled with blurry, noisy, or redundant frames. This is due to frame rate to camera speed ratios that are often higher than necessary; camera and lens imperfections and limitations that result in imaging noise; and occasional jerky motions of the camera that result in motion blur; all of which can significantly affect the performance of the videogrammetric pipeline. To tackle these issues, this paper proposes a novel method for automating the selection of an optimized number of informative, high quality frames. According to this method, as the first step, blurred frames are removed using the thresholds determined based on a minimum level of frame quality required to obtain robust results. Then, an optimum number of key frames are selected from the remaining frames using the selection criteria devised by the authors. Experimental results show that the proposed method outperforms existing methods in terms of improved 3D reconstruction results, while maintaining the optimum number of extracted frames needed to generate high quality 3D point clouds.

Generating Absolute-Scale Point Cloud Data of Built Infrastructure Scenes Using a Monocular Camera Setting

AbstractThe global scale of point cloud data (PCD) generated through monocular photography and videogrammetry is unknown and can be calculated using at least one known dimension of the scene. Measuring one or more dimensions for this purpose induces a manual step in the three-dimensional reconstruction process; this increases the effort and reduces the speed of reconstructing scenes, and induces substantial human error in the process due to the high level of measurement accuracy needed. Other ways of measuring such dimensions are based on acquiring additional information by either using extra sensors or specific classes of objects existing in the scene; it was found that these solutions are not simple, cost effective, or general enough to be considered practical for reconstructing both indoor and outdoor built infrastructure scenes. To address the issue, this paper proposes a novel method for automatically calculating the absolute scale of built infrastructure PCD. A premeasured cube for outdoor scenes an...

Bayesian Nonparametric Adaptive Control Using Gaussian Processes

Most current model reference adaptive control (MRAC) methods rely on parametric adaptive elements, in which the number of parameters of the adaptive element are fixed a priori, often through expert judgment. An example of such an adaptive element is radial basis function networks (RBFNs), with RBF centers preallocated based on the expected operating domain. If the system operates outside of the expected operating domain, this adaptive element can become noneffective in capturing and canceling the uncertainty, thus rendering the adaptive controller only semiglobal in nature. This paper investigates a Gaussian process-based Bayesian MRAC architecture (GP-MRAC), which leverages the power and flexibility of GP Bayesian nonparametric models of uncertainty. The GP-MRAC does not require the centers to be preallocated, can inherently handle measurement noise, and enables MRAC to handle a broader set of uncertainties, including those that are defined as distributions over functions. We use stochastic stability arguments to show that GP-MRAC guarantees good closed-loop performance with no prior domain knowledge of the uncertainty. Online implementable GP inference methods are compared in numerical simulations against RBFN-MRAC with preallocated centers and are shown to provide better tracking and improved long-term learning.