Sergiy Fefilatyev

Horizon Detection Using Machine Learning Techniques

Detecting a horizon in an image is an important part of many image related applications such as detecting ships on the horizon, flight control, and port security. Most of the existing solutions for the problem only use image processing methods to identify a horizon line in an image. This results in good accuracy for many cases and is fast in computation. However, for some images with difficult environmental conditions like a foggy or cloudy sky these image processing methods are inherently inaccurate in identifying the correct horizon. This paper investigates how to detect the horizon line in a set of images using a machine learning approach. The performance of the SVM, J48, and naive Bayes classifiers, used for the problem, has been compared. Accuracy of 90-99% in identifying horizon was achieved on image data set of 20 images

Automatic expression spotting in videos

In this paper, we propose a novel solution for the problem of segmenting macro- and micro-expression frames (or retrieving the expression intervals) in video sequences, which is a prior step for many expression recognition algorithms. The proposed method exploits the non-rigid facial motion that occurs during facial expressions by capturing the optical strain corresponding to the elastic deformation of facial skin tissue. The method is capable of spotting both macro-expressions which are typically associated with expressed emotions and rapid micro- expressions which are typically associated with semi-suppressed macro-expressions. We test our algorithm on several datasets, including a newly released hour-long video with two subjects recorded in a natural setting that includes spontaneous facial expressions. We also report results on a dataset that contains 75 feigned macro-expressions and 37 feigned micro-expressions. We achieve over a 75% true positive rate with a 1% false positive rate for macro-expressions, and a nearly 80% true positive rate for spotting micro-expressions with a .3% false positive rate.

Detection and tracking of marine vehicles in video

This work presents a novel technique for automatic detection and tracking of marine vehicles in video of open sea. The source of video is a video camera mounted on a buoy platform in open sea. Such system is intended to work autonomously, taking video of the surrounding ocean surface and analyzing them on presence of marine vehicles. The proposed technique is based on detection of marine vehicles in individual video frames and tracking the detected targets through the video sequence with the help of a tracking algorithm. Several performance metrics are utilized for performance evaluation of the proposed approach. Accuracy of detection in 90% range is shown on a dataset of 30 short video sequences taken by a prototype of the system.

Toward detection of marine vehicles on horizon from buoy camera

This paper presents a new technique for automatic detection of marine vehicles in open sea from a buoy camera system using computer vision approach. Users of such system include border guards, military, port safety and flow management, sanctuary protection personnel. The system is intended to work autonomously, taking images of the surrounding ocean surface and analyzing them on the subject of presence of marine vehicles. The goal of the system is to detect an approximate window around the ship and prepare the small image for transmission and human evaluation. The proposed computer vision-based algorithm combines horizon detection method with edge detection and post-processing. The dataset of 100 images is used to evaluate the performance of proposed technique. We discuss promising results of ship detection and suggest necessary improvements for achieving better performance.

Tracking Ships from Fast Moving Camera through Image Registration

This paper presents an algorithm that detects and tracks marine vessels in video taken by a nonstationary camera installed on an untethered buoy. The video is characterized by large inter-frame motion of the camera, cluttered background, and presence of compression artifacts. Our approach performs segmentation of ships in individual frames processed with a color-gradient filter. The threshold selection is based on the histogram of the search region. Tracking of ships in a sequence is enabled by registering the horizon images in one coordinate system and by using a multihypothesis framework. Registration step uses an area-based technique to correlate a processed strip of the image over the found horizon line. The results of evaluation of detection, localization, and tracking of the ships show significant increase in performance in comparison to the previously used technique.

Autonomous buoy platform for low-cost visual maritime surveillance: design and initial deployment

We report on the design and evaluation of the initial results of operation of a prototype of an advanced system for maritime security. The system is autonomous and is designed to remain in the ocean for extended periods up to two months. It is based on the Bottom Stationing Ocean Profiler (BSOP), an un-tethered, autonomous platform that stations itself on the sea floor and ascends to the surface at specific time intervals or, potentially, when triggered by certain events such as recognizable acoustic signals, collected and analyzed on board. The surface operations of the system include optical data acquisition, image data analysis, communication with the ground station, and retrieval based functionality. The system is designed to take video and imagery of the surrounding ocean surface and analyze it for the presence of ships, thus, potentially enabling automatic detection and tracking of marine vehicles as they transit in the vicinity of the platform. The system transmits the data to the ground control via bi-directional RF satellite link and can have its mission parameters reprogrammed during the deployment. The described unit is low cost, easy to deploy and recover, and does not reveal itself to the potential targets. The paper describes the system hardware, architecture, algorithms for visual ship detection and tracking.

Active cleaning of label noise

Mislabeled examples in the training data can severely affect the performance of supervised classifiers. In this paper, we present an approach to remove any mislabeled examples in the dataset by selecting suspicious examples as targets for inspection. We show that the large margin and soft margin principles used in support vector machines (SVM) have the characteristic of capturing the mislabeled examples as support vectors. Experimental results on two character recognition datasets show that one-class and two-class SVMs are able to capture around 85% and 99% of label noise examples, respectively, as their support vectors. We propose another new method that iteratively builds two-class SVM classifiers on the non-support vector examples from the training data followed by an expert manually verifying the support vectors based on their classification score to identify any mislabeled examples. We show that this method reduces the number of examples to be reviewed, as well as providing parameter independence of this method, through experimental results on four data sets. So, by (re-)examining the labels of the selective support vectors, most noise can be removed. This can be quite advantageous when rapidly building a labeled data set. HighlightsNovel method for label noise removal from data is introduced.It significantly reduces the required number of examples to be reviewed.Support vectors of SVM classifier can capture around 99% of label noise examples.Two-class SVM captures more label noise examples than one-class SVM classifierCombination of one-class and two-class SVM produces a marginal improvement.

Detection of Anomalous Particles from the Deepwater Horizon Oil Spill Using the SIPPER3 Underwater Imaging Platform

The aim of this study is to investigate a data mining approach to help assess consequences of oil spills in the maritime environment. The approach under investigation is based on detecting suspected oil droplets in the water column adjacent to the Deepwater Horizon oil spill. Our method automatically detects particles in the water, classifies them and provides an interface for visual display. The particles can be plankton, marine snow, oil droplets and more. The focus of this approach is to generalize the methodology utilized for plankton classification using SIPPER (Shadow Imaging Particle Profiler and Evaluation Recorder). In this paper, we report on the application of image processing and machine learning techniques to discern suspected oil droplets from plankton and other particles present in the water. We train the classifier on the data obtained during one of the first research cruises to the site of the Deepwater Horizon oil spill. Suspected oil droplets were visually identified in SIPPER images by an expert. The classification accuracy of the suspected oil droplets is reported and analyzed. Our approach reliably finds oil when it is present. It also classifies some particles (air bubbles and some marine snow), up to 3.3%, as oil in clear water. You can reliably find oil by visually looking at the examples put in the oil class ordered by probability, in which case oil is found in the first 10% of images examined.

Clinical deployment of a medical expert system to increase accruals for clinical trials: Challenges

Before new medical treatments become available to the public, clinicians must conduct extensive trials to determine the efficacy of the novel therapy. In order for the clinical trial to be successful, a significant number of patients with an appropriate set of medical conditions must be accrued. We have implemented a web-based expert system at the H. Lee Moffitt Cancer Center & Research Institute in the Gastrointestinal Tumor Clinic (GITC) to help physicians screen patients for phase II trials. Our system allows physicians to screen a patient for multiple trials simultaneously. Our experiments have shown that adaptation of the system into a clinical environment and the success of the system are related to the amount of time physicians are willing to spend entering data. We also found significant regulatory issues (HIPAA) that make implementation challenging.

Detection of the Vanishing Line of the Ocean Surface from Pairs of Scale-Invariant Keypoints

In this paper, we propose an algorithm for estimating the vanishing line of a stochastically-textured plane in a single image taken by an uncalibrated perspective camera. As an example of such type of texture we take images of ocean surface for which existing methods of vanishing line detection from texture perform poorly. The proposed algorithm relies on finding pairs of similarly looking scale-invariant keypoints that are different in scale. The location of the vanishing line is estimated directly from those pairs of points by finding the vanishing line that represents the consensus of individually found vanishing points. We demonstrate the potential of the proposed method on a number of real images of ocean surface by estimating the horizon line using SIFT keypoints.