Browsing by Subject "Image processing"

Now showing 1 - 10 of 10
  • Thumbnail Image
    Item
    Covariance based point cloud descriptors for object detection and classification
    (2013-08) Fehr, Duc Alexandre
    Processing 3D point data is of primary interest in many areas of computer vision, including object grasping, robot navigation, and 3D object recognition. The recent introduction of cheap range sensors like the Microsoft Kinect has created a great interest in the computer vision community towards developing efficient algorithms for point cloud processing. Previously, in order to capture a point cloud, expensive specialized sensors, such as lasers or dedicated range imagers, were needed; now, range data is readily available from low-cost sensors which provide easily extractable point clouds from a depth map. From here, an interesting challenge is to find different objects in the point cloud. Various descriptors have been introduced to match features in a point cloud. Cheaper sensors are not necessarily designed to produce precise measurements, which entails that the data is not as accurate as a point cloud provided from a laser or a dedicated range finder. There have been feature descriptors that have been shown to be successful in recognizing objects from point clouds. The aim of this thesis is to introduce techniques from other domains, such as image processing, into the field of 3D point cloud processing in order to improve their rendering, recognition, and classification. Covariances have been proven to be very successful in image processing but other domains as well. This work is a first demonstration of the application of covariances in conjunction with 3D point cloud data.
  • Thumbnail Image
    Item
    Development of an Innovative Prototype Lane Departure Warning System
    (Intelligent Transportation Systems Institute, Center for Transportation Studies, University of Minnesota, 2013-03) Yang, Jiann-Shiou
    Development of various techniques such as lane departure warning (LDW) systems can improve traffic safety significantly. An LDW system should be able to detect when the driver is in danger of departing the road and then trigger an alarm to warn the driver early enough to take corrective action. This report presents the development of a new prototype LDW system. It is mainly an image-based approach to find the vehicle's lateral characteristics and then uses that information to establish an operation algorithm to determine whether a warning signal should be issued based on the status of the vehicle deviating from its heading lane. The system developed takes a mixed approach by integrating the Lucas-Kanade (L-K) optical flow and the Hough transform-based lane detection methods in its implementation. The L-K point tracking is used when the lane boundaries cannot be detected, while the lane detection technique is used when they become available. Even though both techniques are used in the system, only one method is activated at any given time because each technique has its own advantages and also disadvantages. The developed LDW system was road tested on I-35, US-53, Rice Lake Road, Martin Road, and Jean Duluth Road. Overall, the system operates correctly as expected, with a false alarm occurring only roughly 1.18% of the operation time. This report presents the system implementation together with findings. Factors that could affect the system performance are also discussed.
  • Thumbnail Image
    Item
    Error Bounds for Finite-difference Methods for Rudin-Osher-Fatemi Image Smoothing
    (University of Minnesota. Institute for Mathematics and Its Applications, 2009-09) Wang, Jinhyue; Lucier, Bradley J.
  • Thumbnail Image
    Item
    Estimation of vehicle lateral position using the optical flow method
    (2013-05) Shrestha, Rini
    With the increasing number of vehicles on the road the number of accidents have also been increasing. Development of various techniques such as lane departure warning systems that helps drivers to assist in driving can help reduce the number of accidents significantly. In this thesis, we attempt to develop such lane departure warning system by estimating the vehicle's lateral position. The lateral position of vehicle can be known if the heading angle of the vehicle can be determined. Therefore, this study focuses on determining heading angle and works toward development of the lane departure warning system based on image processing techniques. An in-vehicle camera is used to capture the images of the road in real time. The system then uses homography on this front - view images of the road to remove the perspective effect and transform the images such that the obtained resulting images represent as if they were observed from the top. The histogram equalization is also applied to the images to increase the global contrast by spreading out the most frequent intensity values. Shi and Tomasi corner detection technique has been used to find significant features (corners) in the images and Lucas - Kanade optical flow to track those corners in following images. The heading angle, and thus the lateral displacement, is determined by relating these tracked corners. As part of the study, a number of road tests were conducted on different roads of Duluth, MN and the findings based on the road tests are discussed.
  • Thumbnail Image
    Item
  • Thumbnail Image
    Item
    Image Enhancement Algorithms to Improve Robustness and Accuracy of Pre-trained Neural Networks for Autonomous Driving
    (2023-01) Joshi, Himanshu
    This research proposes a generalized data-driven approach for improving the pre-training image datasets fed to neural networks(NNs). The algorithms developed and tested in this work could substantially enhance the image bringing out the critical spatial information in the image for better NN performance. This image enhancement technique consists of two main components: image colorization and contrast enhancement. Image colorization is implemented to obtain a color-corrected image from a grayscale image. The traditional global contrast enhancement algorithm is extended to Smoothened Variable Local Dynamic Contrast Improvement (SVL-DCI) to boost local contrasts within an image frame that suffers from under/over-exposed lighting conditions. SVL-DCI algorithm is developed and thoroughly tested in the present thesis that could run in real-time as a pre-training algorithm for NNs. We implemented SVL-DCI on the 3P lab dataset of 2470 images and observed competitive improvement in the performance of the investigated NNs for object recognition and lane detection.
  • Thumbnail Image
    Item
    Object tracking in aerial video of smoky environments
    (University of Minnesota. Institute for Mathematics and Its Applications, 2011-02) Rivera, Mariano; Fernandes, Praphat Xavier; Hernández, Francisco; Martínez, Hugo; McDonald, Matt; Ohlmacher, Scott W.; Wiens, Jeffery
  • Thumbnail Image
    Item
    Structured sparse models with applications
    (2012-10) Sprechmann, Pablo G.
    Sparse models assume minimal prior knowledge about the data, asserting that the signal has many coefficients close or equal to zero when represented in a given domain. From a data modeling point of view, sparsity can be seen as a form of regularization, that is, as a device to restrict or control the set of coefficient values which are allowed in the model to produce an estimate of the data. In this way, flexibility of the model (that is, the ability of a model to fit given data) is reduced, and robustness is gained by ruling out unrealistic estimates of the coefficients. Implicitly, standard sparse models give the same relevance to all of the very large number of subsets of sparse nonzero coefficients (a number which grows exponentially with the number of atoms in the dictionary). This assumption can be easily proved false in many practical cases. Signals have in general a richer underlying structure that is merely disregarded by the model. In many situations, standard sparse models represent a very good trade off between model simplicity and accuracy. However, many practical situations could greatly benefit from exploiting the structure present in the data, potentially for interpretability purposes, improve performance and faster processing. The main goal of this thesis is to explore different ways of including data structure into sparse models and to evaluate them in real image and signal processing applications. The main directions of research are: (i) extending sparse models through imposing structure in the sparsity patterns of non-zero coefficients in order to stabilize the estimation and account for valuable prior knowledge of the signals; (ii) analyzing how this impacts in challenging real applications where the problem of estimating the model coefficients is very ill-posed. As a fundamental example, the problem of monaural source separation will be extensively evaluated throughout the thesis; (iii) studying ways of exploiting the underlying structure of the data in order to speed up the coding process. One of the most important challenges in sparse modeling is the relatively high computational complexity of the inference algorithms, which is of critical importance when dealing with very large scale (modern-size) applications as well as real-time processing.
  • Thumbnail Image
    Item
    Uncertainty in Economic Optimum Nitrogen Rate and Accuracy of Drone Hyperspectral Imaging for Precision Nitrogen Management in Maize
    (2021-06) Nigon, Tyler
    Over the past century, the global nitrogen cycle has been substantially altered by nitrogen fixation via the Haber-Bosch process. This fixed nitrogen is primarily used as fertilizer, ultimately supporting food, fuel, and fiber production for the ever-growing global human population. In the United States, maize production uses far more Haber-Bosch nitrogen than any other activity. Nitrogen fertilizer is necessary to achieve optimal profits, but also contributes to unintended environmental pollution, especially when applied in excess. A great deal of research has been conducted over the past several decades to improve maize nitrogen fertilizer recommendations. However, recommendations are still less accurate than necessary at the field level to successfully balance the resulting economic and environmental tradeoffs. The overarching goal of this research was to improve the understanding and extensibility of precision nitrogen fertilizer recommendations for maize. This goal was addressed by focusing on two areas that currently leads to much of the uncertainty around recommendations: i) uncertainty around the modeled economic optimal nitrogen rate derived from yield response data and ii) quality control standards for developing and implementing remote sensing-based models for predicting in-season crop nitrogen status. The focal point of each of these research areas is the spatial and temporal variation that exists in nitrogen requirements across space and from season to season. The results from this research show there was substantial variability in the modeled economic optimal nitrogen rates for several sites across Minnesota (90% confidence intervals ranged from 42 to 485 kg ha-1). Any regional economic or social analyses are only as reliable as this range of uncertainty around the modeled optimal rate, so caution must be taken to avoid misguided policy recommendations. Hyperspectral imaging was used to accurately predict early-season maize nitrogen uptake (relative RMSE < 24%). Optimizing the image processing protocol improved accuracy further, but it remains a challenge to predict the optimal nitrogen rate from early-season nitrogen status metrics such as nitrogen uptake. Doing so is a necessary step towards estimating nitrogen need and applying nitrogen at the most suitable rates and times so nitrogen recovery is maximized and nutrient loss is minimized.
  • Thumbnail Image
    Item
    Wide Area Detection System (WADS): Image Recognition Algorithms
    (1990-02) Michalopoulos, Panos; Johnston, S. E.; Fundakowski, R. A.; Fitch, R. C.
    Vehicle detection through machine vision is one of the most promising advanced technologies available today for dealing with the problem of urban traffic congestion. In this project an existing Wide Area Detection System (WADS) was improved for performing detection under all weather, traffic, and artifact conditions (e.g. shadows, reflections, lightning, etc. As a result of this and other related research efforts by the same team, a real-time (instead of the initially envisioned off-line) multispot breadboard WADS system was developed, installed, tested, and demonstrated in several real-life situations. The system can simultaneously detect traffic at multiple points within the field of the camera's view and emulates loop detectors. The test results to this point suggest high accuracy levels, comparable to loop detectors, while speed measurement appears to be more accurate than loops. Live demonstrations and off-line presentations generated the enthusiasm and support of practicing engineers and public officials. They also suggest that the WADS system developed in this project is the most advanced one available today. Despite this, further work remains to be done prior to production. This includes extensive field testing and validation as well as implementation of applications possibly through demonstration projects. This report describes the WADS algorithm development and testing and makes recommendations for field implementation of the technology.