Browsing by Author "Papanikolopoulos, Nikolaos"

Now showing 1 - 8 of 8
  • Thumbnail Image
    Item
    Autonomous Mobile Asphalt Density Profiling Robot to Reduce Worker Risk
    (Center for Transportation Studies, University of Minnesota, 2023-06) Morris, Ted; Papanikolopoulos, Nikolaos
    MnDOT pavement construction personnel have lately improved quality assurance (QA) through the use of nondestructive air coupled ground penetrating radar sensors. Although proving to be accurate, the acquisition process can be manually intensive and hazardous especially when deployed adjacent to prevailing traffic. The primary objective of this project was to deliver to MnDOT two low-cost, modular, highly transportable, mobile robot platforms designed specifically for pavement density profile testing. Several field tests were performed to assess feasibility of the platform under different operational scenarios. Modularity was ensured by integrating separate, distributed, plug-and-play modules that could be reused for other mobile platforms, should the need arise for future implementations. By implementing two robots, the transferability of the architecture was demonstrated. The mobile robotic platforms were purposely assembled from widely available, low-cost, commercial, off-the-shelf components to minimize overall cost, recognizing that the landscape for such platforms has been evolving rapidly.
  • Thumbnail Image
    Item
    Computer Aided Diagnosis of Skin Lesions from Morphological Features
    (2018-08-24) Heller, Nicholas; Bussmann, Erika; Shah, Aneri; Dean, Joshua; Papanikolopoulos, Nikolaos
    Skin cancer is the most common cancer, accounting for over 40% of all cancer cases. The morphological features of skin lesions are an integral component of skin cancer detection and diagnosis. With the rapid progress in the field of image classification, increasing attention has been put towards the Computer Aided Diagnosis of skin lesions based on  their  morphological  features.  The  International  Skin  Imaging  Collaboration (ISIC) Archive is the largest publicly available collection of dermoscopic images of skin lesions, and in 2018 ISIC hosted an image recognition challenge for dermoscopic images. This short paper describes the CAD system that we submitted to this challenge.
  • Thumbnail Image
    Item
    Counting Empty Parking Spots at Truck Stops Using Computer Vision
    (Center for Transportation Studies, 2011-05) Pushkar, Modi; Vassilios, Morellas; Papanikolopoulos, Nikolaos
    For at least the past decade, truck driver fatigue has been thought to be a contributing factor in a number of heavy truck accidents. For better utilization of truck stops and to provide truck drivers with safe rest options, we are designing an automated truck stop management system that can compute occupancy rates at stops and notify drivers about the availability of parking spots using variable message displays located about 30 or 40 miles before the stop. Our system detects, classifies and localizes vehicles on the truck stop's grounds by using a set of video cameras, from which video frames are analyzed in real-time.
  • Thumbnail Image
    Item
    Data Mining of Traffic Video Sequences
    (University of Minnesota Center for Transportation Studies, 2009-09) Joshi, Ajay J.; Papanikolopoulos, Nikolaos
    Automatically analyzing video data is extremely important for applications such as monitoring and data collection in transportation scenarios. Machine learning techniques are often employed in order to achieve these goals of mining traffic video to find interesting events. Typically, learning-based methods require significant amount of training data provided via human annotation. For instance, in order to provide training, a user can give the system images of a certain vehicle along with its respective annotation. The system then learns how to identify vehicles in the future - however, such systems usually need large amounts of training data and thereby cumbersome human effort. In this research, we propose a method for active learning in which the system interactively queries the human for annotation on the most informative instances. In this way, learning can be accomplished with lesser user effort without compromising performance. Our system is also efficient computationally, thus being feasible in real data mining tasks for traffic video sequences.
  • Thumbnail Image
    Item
    Deployment of Practical Methods for Counting Bicycle and Pedestrian Use of a Transportation Facility
    (Intelligent Transportation Systems Institute, Center for Transportation Studies, 2012-01) Somasundaram, Guruprasad; Morellas, Vassilios; Papanikolopoulos, Nikolaos
    The classification problem of distinguishing bicycles from pedestrians for traffic counting applications is the objective of this research project. The scenes that are typically involved are bicycle trails, bridges, and bicycle lanes. These locations have heavy traffic of mainly pedestrians and bicyclists. A vision-based system overcomes many of the shortcomings of existing technologies such as loop counters, buried pressure pads, infra-red counters, etc. These methods do not have distinctive profiles for bicycles and pedestrians. Also most of these technologies require expert installation and maintenance. Cameras are inexpensive and abundant and are relatively easy to use, but they tend to be useful as a counting system only when accompanied by powerful algorithms that analyze the images. We employ state-of-the-art algorithms for performing object classification to solve the problem of distinguishing bicyclists from pedestrians. We detail the challenges that are involved in this particular problem, and we propose solutions to address these challenges. We explore common approaches of global image analysis aided by motion information and compare the results with local image analysis in which we attempt to distinguish the individual parts of the composite object. We compare the classification accuracies of both approaches on real data and present detailed discussion on practical deployment factors.
  • Thumbnail Image
    Item
    Monitoring the Use of HOV and HOT Lanes
    (Intelligent Transportation Systems Institute, Center for Transportation Studies, University of Minnesota, 2013-01) Holec, Eric; Somasundaram, Guruprasad; Papanikolopoulos, Nikolaos; Morellas, Vassilios
    This report presents the formulation and implementation of an automated computer vision and machine learning based system for estimation of the occupancy of passenger vehicles in high-occupancy vehicles and highoccupancy toll (HOV/HOT) lanes. We employ a multi-modal approach involving near-infrared images and highresolution color video images in conjunction with strong maximum margin based classifiers such as support vector machines. We attempt to maximize the information that can be extracted from these two types of images by computing different features. Then, we build classifiers for each type of feature which are compared to determine the best feature for each imaging method. Based on the performance of the classifiers we critique the efficacy of the individual approaches as the costs involved are significantly different.
  • Thumbnail Image
    Item
    Sensing for HOV/HOT Lanes Enforcement
    (Minnesota Department of Transportation, 2017-02) Morris, Ted; Morellas, Vassilios; Canelon-Suarez, Dario; Papanikolopoulos, Nikolaos
    The use and creation of combined high-occupancy vehicle/high-occupancy toll (HOV/HOT Lanes) have become more common in urban areas since all types of road users can take advantage of the lane either as a high- occupancy vehicle or opting in to pay a congestion adjusted free. However, to maintain working integrity of the lanes for all users, stepped enforcement to discourage cheating has been needed as more lanes are added. This study evaluated the capability of a novel image sensor device to automate detection of in-vehicle occupants to flag law enforcement of HOV/HOT lane violators. The sensor device synchronously captures three co-registered images, one in the visible spectrum and two others in the infrared bands. The key idea is that the infrared bands can enhance correct occupancy detection through known phenomenological spectral properties of objects and humans residing inside the vehicle. Several experiments were conducted to determine this capability across varied conditions and scenarios to assess detection segmentation algorithms of vehicle passengers and drivers. Although occupancy detection through vehicle glass could be achieved in many cases, improvements must be made to such a detection system to increase robustness and reliability as a law enforcement tool. These improvements were guided by the experimental results, as well as suggested methods for deployment if this or similar technologies were to be deployed in the future.
  • Thumbnail Image
    Item
    Warning Efficacy of Active Versus Passive Warnings for Unsignalized Intersection and Mid-Block Pedestrian Crosswalks
    (Minnesota Department of Transportation, 2009-01) Smith, Thomas J.; Hammond, Curtis; Somasundaram, Guruprasad; Papanikolopoulos, Nikolaos
    This study evaluated the efficacy of active versus passive warnings at uncontrolled pedestrian (ped) crosswalks (Xwalks), by comparing how these two warnings types influenced behavior of drivers approaching such Xwalks. Vehicle-Xwalk interactions were observed at 18 sites with passive, continuously flashing, or ped-activated warnings, yielding 7,305 no ped present and 596 ped present interactions. Vehicle velocities and accelerations were averaged for each interaction. Findings show no significant effect of warning type on overall velocities for either interaction type. With peds present only, for average velocities at successive 5m distances from the Xwalk, a downward trend in velocities from 25 to 5m is observed for passive and active warning sites, but not for pedactivated warning sites. Various lines of evidence point to a number of sources of ambiguity regarding the salience of uncontrolled Xwalk warnings, resulting in behavioral uncertainty by drivers interacting with such warnings. Mixed findings on effects of warning type in this study point to the need for further analysis of this problem area.