Automated sparse 3D point cloud generation of infrastructure using its distinctive visual features

The commercial far-range (>10 m) spatial data collection methods for acquiring infrastructure’s geometric data are not completely automated because of the necessary manual pre- and/or post-processing work. The required amount of human intervention and, in some cases, the high equipment costs associated with these methods impede their adoption by the majority of infrastructure mapping activities. This paper presents an automated stereo vision-based method, as an alternative and inexpensive solution, to producing a sparse Euclidean 3D point cloud of an infrastructure scene utilizing two video streams captured by a set of two calibrated cameras. In this process SURF features are automatically detected and matched between each pair of stereo video frames. 3D coordinates of the matched feature points are then calculated via triangulation. The detected SURF features in two successive video frames are automatically matched and the RANSAC algorithm is used to discard mismatches. The quaternion motion estimation method is then used along with bundle adjustment optimization to register successive point clouds. The method was tested on a database of infrastructure stereo video streams. The validity and statistical significance of the results were evaluated by comparing the spatial distance of randomly selected feature points with their corresponding tape measurements.

Auto-calibration of a camera system using Image Alignment

Calibration of a camera system is a necessary step in any stereo metric process. It correlates all cameras to a common coordinate system by measuring the intrinsic and extrinsic parameters of each camera. Currently, manual calibration of a camera system is the only way to achieve calibration in civil engineering operations that require stereo metric processes (photogrammetry, videogrammetry, vision based asset tracking, etc). This type of calibration however is time-consuming and labor-intensive. Furthermore, in civil engineering operations, camera systems are exposed to open, busy sites. In these conditions, the position of presumably stationary cameras can easily be changed due to external factors such as wind, vibrations or due to an unintentional push/touch from personnel on site. In such cases manual calibration must be repeated. In order to address this issue, several self-calibration algorithms have been proposed. These algorithms use Projective Geometry, Absolute Conic and Kruppa Equations and variations of these to produce processes that achieve calibration. However, most of these methods do not consider all constraints of a camera system such as camera intrinsic constraints, scene constraints, camera motion or varying camera intrinsic properties. This paper presents a novel method that takes all constraints into consideration to auto-calibrate cameras using an image alignment algorithm originally meant for vision based tracking. In this method, image frames are taken from cameras. These frames are used to calculate the fundamental matrix that gives epipolar constraints. Intrinsic and extrinsic properties of cameras are acquired from this calculation. Test results are presented in this paper with recommendations for further improvement.

Innovative Stereo Vision-Based Approach to Generate Dense Depth Map of Transportation Infrastructure

Three-dimensional (3-D) spatial data of a transportation infrastructure contain useful information for civil engineering applications, including as-built documentation, on-site safety enhancements, and progress monitoring. Several techniques have been developed for acquiring 3-D point coordinates of infrastructure, such as laser scanning. Although the method yields accurate results, the high device costs and human effort required render the process infeasible for generic applications in the construction industry. A quick and reliable approach, which is based on the principles of stereo vision, is proposed for generating a depth map of an infrastructure. Initially, two images are captured by two similar stereo cameras at the scene of the infrastructure. A Harris feature detector is used to extract feature points from the first view, and an innovative adaptive window-matching technique is used to compute feature point correspondences in the second view. A robust algorithm computes the nonfeature point correspondences. Thus, the correspondences of all the points in the scene are obtained. After all correspondences have been obtained, the geometric principles of stereo vision are used to generate a dense depth map of the scene. The proposed algorithm has been tested on several data sets, and results illustrate its potential for stereo correspondence and depth map generation.

Machine Vision-Based Infrastructure As-Built Documentation Using Edge Points

The existing machine vision-based 3D reconstruction software programs provide a promising low-cost and in some cases automatic solution for infrastructure as-built documentation. However in several steps of the reconstruction process, they only rely on detecting and matching corner-like features in multiple views of a scene. Therefore, in infrastructure scenes which include uniform materials and poorly textured surfaces, these programs fail with high probabilities due to lack of feature points. Moreover, except few programs that generate dense 3D models through significantly time-consuming algorithms, most of them only provide a sparse reconstruction which does not necessarily include required points such as corners or edges; hence these points have to be manually matched across different views that could make the process considerably laborious. To address these limitations, this paper presents a video-based as-built documentation method that automatically builds detailed 3D maps of a scene by aligning edge points between video frames. Compared to corner-like features, edge points are far more plentiful even in untextured scenes and often carry important semantic associations. The method has been tested for poorly textured infrastructure scenes and the results indicate that a combination of edge and corner-like features would allow dealing with a broader range of scenes.

Sparsity-Inducing Optimization Algorithm for the Extraction of Planar Structures in Noisy Point-Cloud Data

Most of the manual labor needed to create the geometric building information model (BIM) of an existing facility is spent converting raw point cloud data (PCD) to a BIM description. Automating this process would drastically reduce the modeling cost. Surface extraction from PCD is a fundamental step in this process. Compact modeling of redundant points in PCD as a set of planes leads to smaller file size and fast interactive visualization on cheap hardware. Traditional approaches for smooth surface reconstruction do not explicitly model the sparse scene structure or significantly exploit the redundancy. This paper proposes a method based on sparsity-inducing optimization to address the planar surface extraction problem. Through sparse optimization, points in PCD are segmented according to their embedded linear subspaces. Within each segmented part, plane models can be estimated. Experimental results on a typical noisy PCD demonstrate the effectiveness of the algorithm.

Recovering 3D Structure of Poorly Textured Infrastructure Scenes Using Point and Line Features

Most of the existing automated machine vision-based techniques for as-built documentation of civil infrastructure utilize only point features to recover the 3D structure of a scene. However it is often the case in man-made structures that not enough point features can be reliably detected (e.g. buildings and roofs); this can potentially lead to the failure of these techniques. To address the problem, this paper utilizes the prominence of straight lines in infrastructure scenes. It presents a hybrid approach that benefits from both point and line features. A calibrated stereo set of video cameras is used to collect data. Point and line features are then detected and matched across video frames. Finally, the 3D structure of the scene is recovered by finding 3D coordinates of the matched features. The proposed approach has been tested on realistic outdoor environments and preliminary results indicate its capability to deal with a variety of scenes.

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.© 2012 Elsevier Ltd. All rights reserved.

Innovative Stereo Vision-Based Approach to Generate Dense Depth Map of Transportation Infrastructure

Three-dimensional (3-D) spatial data of a transportation infrastructure contain useful information for civil engineering applications, including as-built documentation, on-site safety enhancements, and progress monitoring. Several techniques have been developed for acquiring 3-D point coordinates of infrastructure, such as laser scanning. Although the method yields accurate results, the high device costs and human effort required render the process infeasible for generic applications in the construction industry. A quick and reliable approach, which is based on the principles of stereo vision, is proposed for generating a depth map of an infrastructure. Initially, two images are captured by two similar stereo cameras at the scene of the infrastructure. A Harris feature detector is used to extract feature points from the first view, and an innovative adaptive window-matching technique is used to compute feature point correspondences in the second view. A robust algorithm computes the nonfeature point correspondences. Thus, the correspondences of all the points in the scene are obtained. After all correspondences have been obtained, the geometric principles of stereo vision are used to generate a dense depth map of the scene. The proposed algorithm has been tested on several data sets, and results illustrate its potential for stereo correspondence and depth map generation.

A Structured Model of Video Reproduces Primary Visual Cortical Organisation

The visual system must learn to infer the presence of objects and features in the world from the images it encounters, and as such it must, either implicitly or explicitly, model the way these elements interact to create the image. Do the response properties of cells in the mammalian visual system reflect this constraint? To address this question, we constructed a probabilistic model in which the identity and attributes of simple visual elements were represented explicitly and learnt the parameters of this model from unparsed, natural video sequences. After learning, the behaviour and grouping of variables in the probabilistic model corresponded closely to functional and anatomical properties of simple and complex cells in the primary visual cortex (V1). In particular, feature identity variables were activated in a way that resembled the activity of complex cells, while feature attribute variables responded much like simple cells. Furthermore, the grouping of the attributes within the model closely parallelled the reported anatomical grouping of simple cells in cat V1. Thus, this generative model makes explicit an interpretation of complex and simple cells as elements in the segmentation of a visual scene into basic independent features, along with a parametrisation of their moment-by-moment appearances. We speculate that such a segmentation may form the initial stage of a hierarchical system that progressively separates the identity and appearance of more articulated visual elements, culminating in view-invariant object recognition.

Modeling natural sounds with modulation cascade processes

Natural sounds are structured on many time-scales. A typical segment of speech, for example, contains features that span four orders of magnitude: Sentences ($\sim1$s); phonemes ($\sim10$−$1$ s); glottal pulses ($\sim 10$−$2$s); and formants ($\sim 10$−$3$s). The auditory system uses information from each of these time-scales to solve complicated tasks such as auditory scene analysis [1]. One route toward understanding how auditory processing accomplishes this analysis is to build neuroscience-inspired algorithms which solve similar tasks and to compare the properties of these algorithms with properties of auditory processing. There is however a discord: Current machine-audition algorithms largely concentrate on the shorter time-scale structures in sounds, and the longer structures are ignored. The reason for this is two-fold. Firstly, it is a difficult technical problem to construct an algorithm that utilises both sorts of information. Secondly, it is computationally demanding to simultaneously process data both at high resolution (to extract short temporal information) and for long duration (to extract long temporal information). The contribution of this work is to develop a new statistical model for natural sounds that captures structure across a wide range of time-scales, and to provide efficient learning and inference algorithms. We demonstrate the success of this approach on a missing data task.

Probabilistic Amplitude Demodulation

Auditory scene analysis is extremely challenging. One approach, perhaps that adopted by the brain, is to shape useful representations of sounds on prior knowledge about their statistical structure. For example, sounds with harmonic sections are common and so time-frequency representations are efficient. Most current representations concentrate on the shorter components. Here, we propose representations for structures on longer time-scales, like the phonemes and sentences of speech. We decompose a sound into a product of processes, each with its own characteristic time-scale. This demodulation cascade relates to classical amplitude demodulation, but traditional algorithms fail to realise the representation fully. A new approach, probabilistic amplitude demodulation, is shown to out-perform the established methods, and to easily extend to representation of a full demodulation cascade.

Independence is elusive: set size effects on encoding precision in visual search.

Looking for a target in a visual scene becomes more difficult as the number of stimuli increases. In a signal detection theory view, this is due to the cumulative effect of noise in the encoding of the distractors, and potentially on top of that, to an increase of the noise (i.e., a decrease of precision) per stimulus with set size, reflecting divided attention. It has long been argued that human visual search behavior can be accounted for by the first factor alone. While such an account seems to be adequate for search tasks in which all distractors have the same, known feature value (i.e., are maximally predictable), we recently found a clear effect of set size on encoding precision when distractors are drawn from a uniform distribution (i.e., when they are maximally unpredictable). Here we interpolate between these two extreme cases to examine which of both conclusions holds more generally as distractor statistics are varied. In one experiment, we vary the level of distractor heterogeneity; in another we dissociate distractor homogeneity from predictability. In all conditions in both experiments, we found a strong decrease of precision with increasing set size, suggesting that precision being independent of set size is the exception rather than the rule.

Atmospheric turbulence mitigation using complex wavelet-based fusion

Restoring a scene distorted by atmospheric turbulence is a challenging problem in video surveillance. The effect, caused by random, spatially varying, perturbations, makes a model-based solution difficult and in most cases, impractical. In this paper, we propose a novel method for mitigating the effects of atmospheric distortion on observed images, particularly airborne turbulence which can severely degrade a region of interest (ROI). In order to extract accurate detail about objects behind the distorting layer, a simple and efficient frame selection method is proposed to select informative ROIs only from good-quality frames. The ROIs in each frame are then registered to further reduce offsets and distortions. We solve the space-varying distortion problem using region-level fusion based on the dual tree complex wavelet transform. Finally, contrast enhancement is applied. We further propose a learning-based metric specifically for image quality assessment in the presence of atmospheric distortion. This is capable of estimating quality in both full-and no-reference scenarios. The proposed method is shown to significantly outperform existing methods, providing enhanced situational awareness in a range of surveillance scenarios. © 1992-2012 IEEE.

Unconstrained monocular 3D human pose estimation by action detection and cross-modality regression forest

This work addresses the challenging problem of unconstrained 3D human pose estimation (HPE) from a novel perspective. Existing approaches struggle to operate in realistic applications, mainly due to their scene-dependent priors, such as background segmentation and multi-camera network, which restrict their use in unconstrained environments. We therfore present a framework which applies action detection and 2D pose estimation techniques to infer 3D poses in an unconstrained video. Action detection offers spatiotemporal priors to 3D human pose estimation by both recognising and localising actions in space-time. Instead of holistic features, e.g. silhouettes, we leverage the flexibility of deformable part model to detect 2D body parts as a feature to estimate 3D poses. A new unconstrained pose dataset has been collected to justify the feasibility of our method, which demonstrated promising results, significantly outperforming the relevant state-of-the-arts. © 2013 IEEE.

A dedicated binding mechanism for the visual control of movement.

The human motor system is remarkably proficient in the online control of visually guided movements, adjusting to changes in the visual scene within 100 ms [1-3]. This is achieved through a set of highly automatic processes [4] translating visual information into representations suitable for motor control [5, 6]. For this to be accomplished, visual information pertaining to target and hand need to be identified and linked to the appropriate internal representations during the movement. Meanwhile, other visual information must be filtered out, which is especially demanding in visually cluttered natural environments. If selection of relevant sensory information for online control was achieved by visual attention, its limited capacity [7] would substantially constrain the efficiency of visuomotor feedback control. Here we demonstrate that both exogenously and endogenously cued attention facilitate the processing of visual target information [8], but not of visual hand information. Moreover, distracting visual information is more efficiently filtered out during the extraction of hand compared to target information. Our results therefore suggest the existence of a dedicated visuomotor binding mechanism that links the hand representation in visual and motor systems.