Robert August Kaucic

A unified framework for tracking through occlusions and across sensor gaps

A common difficulty encountered in tracking applications is how to track an object that becomes totally occluded, possibly for a significant period of time. Another problem is how to associate objects, or tracklets, across non-overlapping cameras, or between observations of a moving sensor that switches fields of regard. A third problem is how to update appearance models for tracked objects over time. As opposed to using a comprehensive multi-object tracker that must simultaneously deal with these tracking challenges, we present a novel, modular framework that handles each of these problems in a unified manner by the initialization, tracking, and linking of high-confidence tracklets. In this track/suspend/match paradigm, we first analyze the scene to identify areas where tracked objects are likely to become occluded. Tracking is then suspended on occluded objects and re-initiated when they emerge from behind the occlusion. We then associate, or match, suspended tracklets with the new tracklets using full kinematic models for object motion and Gibbsian distributions for object appearance in order to complete the track through the occlusion. Sensor gaps are handled in a similar manner, where tracking is suspended when the sensor looks away and then re-initiated when the sensor returns. Changes in object appearance and orientation during tracking are also seamlessly handled in this framework. Tracklets with low lock scores are terminated. Tracking then resumes on untracked movers with corresponding updated appearance models. These new tracklets are then linked back to the terminated ones as appropriate. Fully automatic tracking results from a moving sensor are presented.

Model-based detection of lung nodules in computed tomography exams1

Abstract Rationale and objectives In this study, we developed a prototype model-based computer aided detection (CAD) system designed to automatically detect both solid and subsolid pulmonary nodules in computed tomography (CT) images. By using this CAD algorithm, along with the radiologist’s initial interpretation, we aim to improve the sensitivity of radiologic readings of CT lung exams. Materials and methods We have developed a model-based CAD algorithm through the use of precise mathematic models that capture scanner physics and anatomic information. Our model-based CAD algorithm uses multiple segmentation algorithms to extract noteworthy structures in the lungs and a Bayesian statistical model selection framework to determine the probability of various anatomical events throughout the lung. We tested this algorithm on 50 low-dose CT lung cancer screening cases in which ground truth was produced through readings by three expert chest radiologists. Results Using this model-based CAD algorithm on 50 low-dose CT cases, we measured potential sensitivity improvements of 7% and 5% in two radiologists with respect to all noncalcified nodules, solid and subsolid, greater than 5 mm in diameter. The third radiologist did not miss any nodules in the ground truth set. The CAD algorithm produced 8.3 false positives per case. Conclusion Our prototype CAD system demonstrates promising results as a tool to improve the quality of radiologic readings by increasing radiologist sensitivity. A significant advantage of this model-based approach is that it can be easily extended to support additional anatomic models as clinical understanding and scanning practices improve.

A common set of perceptual observables for grouping, figure-ground discrimination, and texture classification

We present a complete set of perceptual observables that provides a unified image description for grouping, figure-ground separation, and texture analysis. Although much progress has been made recently in treating contours and texture simultaneously for image segmentation and grouping, current approaches rely on different models for contours, regions, and texture such as one-dimensional intensity discontinuities for contours and filter bank responses for texture. This results in expensive computation that arbitrates between these disparate representations at each pixel. In our approach, salient image content such as contours, regions, and texture are represented in a common, low-level framework of image observables. We model the image as a partition of surfaces bounded by intensity discontinuities and derive perceptual measures as relations between neighboring surfaces. This enables us to extend the traditional Gestalt measures based on local edge geometry and contrast to region-based measures that jointly exploit large scale image topology, photometry, and geometry. These measures provide a natural basis for grouping on multidimensional similarity criteria and texture is directly derived as relational properties on local region neighborhoods. The viability of our model is demonstrated by applying the common observables to texture recognition, figure-ground separation, and generic image segmentation.

Multimodal Industrial Inspection and Analysis

Diagnosis of complex engineering systems requires the use of multiple sensor sources to acquire information. In this paper we present a survey of multimodal data acquisition systems for nondestructive testing (NDT) and engineering analysis. We begin with a summary of the relative strengths and weaknesses of individual NDT modalities. Thereafter we present existing multimodal inspection hardware systems that use complementary NDT sensors. The advantages of such multimodal data acquisition over conventional single modality sensors in inspection and analysis are highlighted. Possible approaches to fuse complementary multimodal sensor data are discussed. We conclude with possible directions for the future development of multimodal inspection systems.

Learning-based automatic defect recognition with computed tomographic imaging

The use of image-based automatic defect recognition (ADR) systems in a production line often requires strict processing-time specifications. On the other hand, the typical high-performance requirement of such system calls for the use of sophisticated, computationally-complex algorithms. Addressing the conflicting requirements of fast throughput and high detection performance is a significant challenge. In this paper we present a 3D learning-based ADR approach for industrial parts. The proposed method first extracts defect candidate regions using morphological closing and template matching. Then a local registration-based approach is utilized to produce accurate defect segmentation mask. Finally, 29 features including geometric features and texture features derived from grey level co-occurrence matrix are calculated for each candidate region, and a fast random forests classifier is used to classify the candidate regions as defect or defect-free. This approach was developed into a fully automated system for detecting casting defects in aluminum industrial parts depicted in 3D Computed Tomographic (CT) images. The system was tested on 31 images with 49 cavities and porosities defects, achieving a sensitivity of 94% with an average 3.5 false detections per part.

Single View Metrology: A Practical Example

In machine vision applications it is often desirable to make metric measurements with a single camera. However, operational constraints can limit the use of off-line calibration methods and the lack of reference planes can prevent the use of current single-view metrology techniques. In this paper, we describe the design, implementation, and validation of a visual inspection system for the measurement of angles from a single, uncalibrated camera. We provide a step-by-step, single-view metrology approach which we demonstrate in an operational setting. We first assess the imaging conditions present using hypothesis testing based on a norm-induced metric between possible world to image mappings. Our metric provides a measure of localization error derived from physical measures in the world. The hypothesis testing demonstrates that affine viewing conditions are present in our setting and we show how the affine approximation can be used to make measurements of points off the plane of interest. We employ our algorithm in a manufacturing plant to make blade angle measurements of variable stator vanes in jet engines.

Automatic Deformation Detection for Aircraft Engine Disk Inspection

Computer vision algorithms are seeing increased use in industrial inspection applications. Here, we present an “Aid to Visual” system that can detect post deformations of less than 0.005 inches in jet engine high pressure turbine disks. We create a gold-standard reference post from the posts of sample turbine disks and then use registration, edge detection, and curve-similarity algorithms to identify unacceptable post deformations. We address the challenges associated with adapting academic algorithms for use in functioning inspection systems. We present novel solutions to deal with practical issues such as accuracy, speed, robustness, and ease of use. We also present a novel, highly-efficient sub-pixel contour matching algorithm and demonstrate the effectiveness of using sub-pixel distance calculation. We demonstrate overall error rates less than 1% on over 2400 images of posts. We have integrated our algorithms into the commercial LabVIEW software running on the Aid To Visual workstation. Our algorithms will enable plant-factory inspectors to identify minute post deformations that were previously difficult to detect.

Model-based detection of lung lesions in CT exams

Abstract The thorough detection of nodules in high-resolution CT lung scans is an increasingly difficult, labor-intensive, and critical radiological task. Recent clinical research on early lung cancer CT presentation has demonstrated the significant clinical need to detect the more subtle subsolid nodules as well as the traditional solid nodule. We have developed a model-based computer-aided detection (CAD) algorithm designed to automatically detect both of these nodule presentation types through the use of precise mathematical models that capture scanner physics and anatomy and pathology domain knowledge. Our model-based CAD algorithm utilizes a Bayesian framework for determining the probability of multiple competing anatomical and pathological events throughout the lung. Using this model-based CAD algorithm on 50 low-dose CT lung cancer screening cases, we measured a 3.9% average improvement in radiologist sensitivity (93.8% to 97.7%) with 8.3 false positives per case for all nodules ≥5 mm in size. This model-based approach can be easily extended to support additional anatomy and pathology models as clinical understanding and scanning practices improve.