Riccardo Gherardi

Structure-and-motion pipeline on a hierarchical cluster tree

This papers introduces a novel hierarchical scheme for computing Structure and Motion. The images are organized into a tree with agglomerative clustering, using a measure of overlap as the distance. The reconstruction follows this tree from the leaves to the root. As a result, the problem is broken into smaller instances, which are then separately solved and combined. Compared to the standard sequential approach, this framework has a lower computational complexity, it is independent from the initial pair of views, and copes better with drift problems. A formal complexity analysis and some experimental results support these claims.

Improving the efficiency of hierarchical structure-and-motion

We present a completely automated Structure and Motionpipeline capable of working with uncalibrated images with varying internal parameters and no ancillary information. The system is based on a novel hierarchical scheme which reduces the total complexity by one order of magnitude. We assess the quality of our approach analytically by comparing the recovered point clouds with laser scans, which serves as ground truth data.

Hierarchical structure-and-motion recovery from uncalibrated images

We describe a hierarchical structure-from-motion pipeline.No information is needed beside images themselves.The pipeline proved successful in real-world tasks. This paper addresses the structure-and-motion problem, that requires to find camera motion and 3D structure from point matches. A new pipeline, dubbed Samantha, is presented, that departs from the prevailing sequential paradigm and embraces instead a hierarchical approach. This method has several advantages, like a provably lower computational complexity, which is necessary to achieve true scalability, and better error containment, leading to more stability and less drift. Moreover, a practical autocalibration procedure allows to process images without ancillary information. Experiments with real data assess the accuracy and the computational efficiency of the method.

Automatic selection of MRF control parameters by reactive tabu search

This paper presents an optimization technique to automatically select a set of control parameters for a Markov random field. The method is based on the reactive tabu search strategy, and requires to define a suitable fitness function that measures the performance of the MRF algorithm with a given parameters set. The technique is applied to stereo matching thanks to the availability of ground truth disparity maps. Experiments with synthetic and real images illustrate the approach.

Visual change detection on tunnel linings

We describe an automated system for detecting, localising, clustering and ranking visual changes on tunnel surfaces. The system is designed to provide assistance to expert human inspectors carrying out structural health monitoring and maintenance on ageing tunnel networks. A three-dimensional tunnel surface model is first recovered from a set of reference images using Structure from Motion techniques. New images are localised accurately within the model and changes are detected versus the reference images and model geometry. We formulate the problem of detecting changes probabilistically and evaluate the use of different feature maps and a novel geometric prior to achieve invariance to noise and nuisance sources such as parallax and lighting changes. A clustering and ranking method is proposed which efficiently presents detected changes and further improves the inspection efficiency. System performance is assessed on a real data set collected using a low-cost prototype capture device and labelled with ground truth. Results demonstrate that our system is a step towards higher frequency visual inspection at a reduced cost.

Confidence-based cost modulation for stereo matching

We present a novel operator to be applied at raw matching costs in the context of low level vision tasks such as stereo matching or optical flow. It aims at improving matching reliability by efficiently modulating pixel-wise pairing costs, injecting a confidence backed bias before the aggregation step. It works analyzing a noisy estimate of the correspondances in order to favor or prune potential matches. We test the operator by developing a local, realtime stereo matching algorithm and showing that our solution can drastically clean the resulting depth map while also reducing border bleeding. Its good performance is also evaluated quantitavely by testing the algorithm against the popular Middlebury benchmark where our local greedy implementation is able to obtain results comparable to those of naive global approaches.

Efficient Visualization of Architectural Models from a Structure and Motion Pipeline

State of the art three dimensional reconstruction pipelines can nowadays produce models up to several million polygons without any human intervention from a set of digital images or vide o. Such models are able to stretch the rendering capabilities of current hardware. We propose to augment a typical structure from motion pipeline with two additional steps, automatic fitting of high-level solid primitives and relief ma ps extraction, thus recovering both the overall structure of a building and its fine geometry. This not only g birth to a more tractable and semantic model of the imaged scene, but allows for efficient and compellin g rendering. We substantiate our claims showing a complete example of the described system.

Detecting Change for Multi-View, Long-Term Surface Inspection.

We describe a system for the detection of changes in multiple views of a tunnel surface. From data gathered by a robotic inspection rig, we use a structure-from-motion pipeline to build panoramas of the surface and register images from different time instances. Reliably detecting changes such as hairline cracks, water ingress and other surface damage between the registered images is a challenging problem: achieving the best possible performance for a given set of data requires sub-pixel precision and careful modelling of the noise sources. The task is further complicated by factors such as unavoidable registration error and changes in image sensors, capture settings and lighting. Our contribution is a novel approach to change detection using a two-channel convolutional neural network. The network accepts pairs of approximately registered image patches taken at different times and classifies them to detect anomalous changes. To train the network, we take advantage of synthetically generated training examples and the homogeneity of the tunnel surfaces to eliminate most of the manual labelling effort. We evaluate our method on field data gathered from a live tunnel over several months, demonstrating it to outperform existing approaches from recent literature and industrial practice.

Samantha: Towards Automatic Image-Based Model Acquisition

In this paper we describe SAMANTHA, a Structure and Motion pipeline from images which is both more robust and computationally cheaper than current competing approaches. Pictures are organized into a hierarchical tree which has single images as leaves and partial reconstructions as internal nodes. The method proceeds bottom up until it reaches the root node, corresponding to the final result. This framework is one order of magnitude faster than sequential approaches, inherently parallel, less sensitive to the error accumulation causing drift and truly uncalibrated, not needing EXIF metadata to be present in pictures. We have verified the quality of our reconstructions both qualitatively producing compelling point clouds and quantitatively, comparing them with laser scans serving as ground truth. We also show how to automatically extract a meaningful collection of planar patches obtaining a compact, stable representation of scenes.

Optimal parameter estimation for MRF stereo matching

This paper presents an optimisation technique to select automatically a set of control parameters for a Markov Random Field applied to stereo matching. The method is based on the Reactive Tabu Search strategy, and requires to define a suitable fitness function that measures the performance of the MRF stereo algorithm with a given parameters set. This approach have been made possible by the recent availability of ground-truth disparity maps. Experiments with synthetic and real images illustrate the approach.