Christoph Strecha

BRIEF: binary robust independent elementary features

We propose to use binary strings as an efficient feature point descriptor, which we call BRIEF. We show that it is highly discriminative even when using relatively few bits and can be computed using simple intensity difference tests. Furthermore, the descriptor similarity can be evaluated using the Hamming distance, which is very efficient to compute, instead of the L2 norm as is usually done. As a result, BRIEF is very fast both to build and to match. We compare it against SURF and U-SURF on standard benchmarks and show that it yields a similar or better recognition performance, while running in a fraction of the time required by either.

On benchmarking camera calibration and multi-view stereo for high resolution imagery

In this paper we want to start the discussion on whether image based 3D modelling techniques can possibly be used to replace LIDAR systems for outdoor 3D data acquisition. Two main issues have to be addressed in this context: (i) camera calibration (internal and external) and (ii) dense multi-view stereo. To investigate both, we have acquired test data from outdoor scenes both with LIDAR and cameras. Using the LIDAR data as reference we estimated the ground-truth for several scenes. Evaluation sets are prepared to evaluate different aspects of 3D model building. These are: (i) pose estimation and multi-view stereo with known internal camera parameters; (ii) camera calibration and multi-view stereo with the raw images as the only input and (iii) multi-view stereo.

BRIEF: Computing a Local Binary Descriptor Very Fast

Binary descriptors are becoming increasingly popular as a means to compare feature points very fast while requiring comparatively small amounts of memory. The typical approach to creating them is to first compute floating-point ones, using an algorithm such as SIFT, and then to binarize them. In this paper, we show that we can directly compute a binary descriptor, which we call BRIEF, on the basis of simple intensity difference tests. As a result, BRIEF is very fast both to build and to match. We compare it against SURF and SIFT on standard benchmarks and show that it yields comparable recognition accuracy, while running in an almost vanishing fraction of the time required by either.

LDAHash: Improved Matching with Smaller Descriptors

SIFT-like local feature descriptors are ubiquitously employed in computer vision applications such as content-based retrieval, video analysis, copy detection, object recognition, photo tourism, and 3D reconstruction. Feature descriptors can be designed to be invariant to certain classes of photometric and geometric transformations, in particular, affine and intensity scale transformations. However, real transformations that an image can undergo can only be approximately modeled in this way, and thus most descriptors are only approximately invariant in practice. Second, descriptors are usually high dimensional (e.g., SIFT is represented as a 128-dimensional vector). In large-scale retrieval and matching problems, this can pose challenges in storing and retrieving descriptor data. We map the descriptor vectors into the Hamming space in which the Hamming metric is used to compare the resulting representations. This way, we reduce the size of the descriptors by representing them as short binary strings and learn descriptor invariance from examples. We show extensive experimental validation, demonstrating the advantage of the proposed approach.

Combined Depth and Outlier Estimation in Multi-View Stereo

In this paper, we present a generative model based approach to solve the multi-view stereo problem. The input images are considered to be generated by either one of two processes: (i) an inlier process, which generates the pixels which are visible from the reference camera and which obey the constant brightness assumption, and (ii) an outlier process which generates all other pixels. Depth and visibility are jointly modelled as a hiddenMarkov Random Field, and the spatial correlations of both are explicitly accounted for. Inference is made tractable by an EM-algorithm, which alternates between estimation of visibility and depth, and optimisation of model parameters. We describe and compare two implementations of the E-step of the algorithm, which correspond to the Mean Field and Bethe approximations of the free energy. The approach is validated by experiments on challenging real-world scenes, of which two are contaminated by independently moving objects.

Making Background Subtraction Robust to Sudden Illumination Changes

Modern background subtraction techniques can handle gradual illumination changes but can easily be confused by rapid ones. We propose a technique that overcomes this limitation by relying on a statistical model, not of the pixel intensities, but of the illumination effects. Because they tend to affect whole areas of the image as opposed to individual pixels, low-dimensional models are appropriate for this purpose and make our method extremely robust to illumination changes, whether slow or fast.

Efficient large-scale multi-view stereo for ultra high-resolution image sets

We present a new approach for large-scale multi-view stereo matching, which is designed to operate on ultra high-resolution image sets and efficiently compute dense 3D point clouds. We show that, using a robust descriptor for matching purposes and high-resolution images, we can skip the computationally expensive steps that other algorithms require. As a result, our method has low memory requirements and low computational complexity while producing 3D point clouds containing virtually no outliers. This makes it exceedingly suitable for large-scale reconstruction. The core of our algorithm is the dense matching of image pairs using DAISY descriptors, implemented so as to eliminate redundancies and optimize memory access. We use a variety of challenging data sets to validate and compare our results against other algorithms.

Dynamic and scalable large scale image reconstruction

Recent approaches to reconstructing city-sized areas from large image collections usually process them all at once and only produce disconnected descriptions of image subsets, which typically correspond to major landmarks. In contrast, we propose a framework that lets us take advantage of the available meta-data to build a single, consistent description from these potentially disconnected descriptions. Furthermore, this description can be incrementally updated and enriched as new images become available. We demonstrate the power of our approach by building large-scale reconstructions using images of Lausanne and Prague.

Surface Construction by a 2-D Differentiation-Integration Process: A Neurocomputational Model for Perceived Border Ownership, Depth, and Lightness in Kanizsa Figures

Human visual perception is a fundamentally relational process: Lightness perception depends on luminance ratios, and depth perception depends on occlusion (difference of depth) cues. Neurons in low-level visual cortex are sensitive to the difference (but not the value itself) of signals, and these differences have to be used to reconstruct the input. This process can be regarded as a 2-dimensional differentiation and integration process: First, differentiated signals for depth and lightness are created at an earlier stage of visual processing and then 2-dimensionally integrated at a later stage to construct surfaces. The subjective filling in of physically missing parts of input images (completion) can be explained as a property that emerges from this surface construction process. This approach is implemented in a computational model, called DISC (Differentiation-Integration for Surface Completion). In the DISC model, border ownership (the depth order at borderlines) is computed based on local occlusion cues (L- and T-junctions) and the distribution of borderlines. Two-dimensional integration is then applied to construct surfaces in the depth domain, and lightness values are in turn modified by these depth measurements. Illusory percepts emerge through the surface-construction process with the development of illusory border ownership and through the interaction between depth and lightness perception. The DISC model not only produces a central surface with the correctly modified lightness values of the original Kanizsa figure but also responds to variations of this figure such that it can distinguish between illusory and nonillusory configurations in a manner that is consistent with human perception.

A Probabilistic Approach to Large Displacement Optical Flow and Occlusion Detection

This paper deals with the computation of optical flow and occlusion detection in the case of large displacements. We propose a Bayesian approach to the optical flow problem and solve it by means of differential techniques. The images are regarded as noisy measurements of an underlying ’true’ image-function. Additionally, the image data is considered incomplete, in the sense that we do not know which pixels from a particular image are occluded in the other images. We describe an EM-algorithm, which iterates between estimating values for all hidden quantities, and optimizing the current optical flow estimates by differential techniques. The Bayesian way of describing the problem leads to more insight in existing differential approaches, and offers some natural extensions to them. The resulting system involves less parameters and gives an interpretation to the remaining ones. An important new feature is the photometric detection of occluded pixels. We compare the algorithm with existing optical flow methods on ground truth data. The comparison shows that our algorithm generates the most accurate optical flow estimates. We further illustrate the approach with some challenging real-world examples.