This paper presents a method for extracting distinctive invariant features from images that can be used to perform reliable matching between different views of an object or scene. The features are invariant to image scale and rotation, and are shown to provide robust matching across a substantial range of affine distortion, change in 3D viewpoint, addition of noise, and change in illumination. The features are highly distinctive, in the sense that a single feature can be correctly matched with high probability against a large database of features from many images. This paper also describes an approach to using these features for object recognition. The recognition proceeds by matching individual features to a database of features from known objects using a fast nearest-neighbor algorithm, followed by a Hough transform to identify clusters belonging to a single object, and finally performing verification through least-squares solution for consistent pose parameters. This approach to recognition can robustly identify objects among clutter and occlusion while achieving near real-time performance.

/pdf/distinctive-image-features-from-scale-invariant-keypoints-3waurjqzke.pdf

Distinctive Image Features from Scale-Invariant Keypoints

We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

/pdf/going-deeper-with-convolutions-1yobw2o2ds.pdf

Going deeper with convolutions

Downloading the book in this website lists can give you more advantages. It will show you the best book collections and completed collections. So many books can be found in this website. So, this is not only this multiple view geometry in computer vision. However, this book is referred to read because it is an inspiring book to give you more chance to get experiences and also thoughts. This is simple, read the soft file of the book and you get it.

Multiple View Geometry in Computer Vision.

We propose a flexible technique to easily calibrate a camera. It only requires the camera to observe a planar pattern shown at a few (at least two) different orientations. Either the camera or the planar pattern can be freely moved. The motion need not be known. Radial lens distortion is modeled. The proposed procedure consists of a closed-form solution, followed by a nonlinear refinement based on the maximum likelihood criterion. Both computer simulation and real data have been used to test the proposed technique and very good results have been obtained. Compared with classical techniques which use expensive equipment such as two or three orthogonal planes, the proposed technique is easy to use and flexible. It advances 3D computer vision one more step from laboratory environments to real world use.

https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/tr98-71.pdf

A flexible new technique for camera calibration

Stereo matching is one of the most active research areas in computer vision. While a large number of algorithms for stereo correspondence have been developed, relatively little work has been done on characterizing their performance. In this paper, we present a taxonomy of dense, two-frame stereo methods designed to assess the different components and design decisions made in individual stereo algorithms. Using this taxonomy, we compare existing stereo methods and present experiments evaluating the performance of many different variants. In order to establish a common software platform and a collection of data sets for easy evaluation, we have designed a stand-alone, flexible C++ implementation that enables the evaluation of individual components and that can be easily extended to include new algorithms. We have also produced several new multiframe stereo data sets with ground truth, and are making both the code and data sets available on the Web.

/pdf/a-taxonomy-and-evaluation-of-dense-two-frame-stereo-4c20etkubp.pdf

A taxonomy and evaluation of dense two-frame stereo correspondence algorithms

A digital multiplier for multiplying together W-bit digit-serial multiplier and multiplicand signals includes a combinational array of multiplier cells arranged in N rows and W columns. A digit-serial-in/parallel-out register supplies respective bits of each successive multiplicand signal to the W columns of the array, the N rows of which receive respective bits of each successive digit of the multiplier signal. After each earlier digit of the multiplier is processed, the carry and sum bits are forwarded without column shift and with one column shift, respectively, from the final row to the first row of multiplier cells. This scrolls the operation of the W-column-by-N-row multiplier cell array, allowing it to be used M times for each word of the multiplier signal, one for each of the M digits in a W-bit word of the multiplier signal. The sum bits from the final column of multiplier cells provide the minor product output signal of the digital multiplier in digit-serial format, as each digit of the multiplier signal is processed. In processing the final digit of the multiplier, the carry and sum bits are shifted differentially by one bit place and are conveyed by shift registers to a digit-serial final adder, the full sum output of which supplies the major product output signal of the digital multiplier in digit-serial form. Supplying the minor and major product output signals in parallel permits pipelined operation of the apparatus.

Bit-sliced digit-serial multiplier

Automatic inspection of pipelines has great potential to increase the efficiency and objectivity of pipeline inspection. 3-D pipeline reconstruction can provide a valuable addition to automatic pipeline inspection. We present a system which can reconstruct the inner surface of buried pipelines from multiple fisheye images captured inside the pipes. The reconstruction output will facilitate pipe condition analysis. Whereas the pipelines are huge, a fatal defect is usually as tiny as a fine crack. Therefore efficiency and accuracy both become the key requisites of our system. The repetitive patterns on the pipe surface and the poor illumination condition during photographing further increase the difficulty. We combine several successful methods in the literature as well as new methods proposed by ourselves and finally build up a system, which can reconstruct pipe surface not only accurately but also quickly. The system has been tested on real pipe images and shows promising performance.

/pdf/pipeline-reconstruction-from-fisheye-images-3q1l0trne9.pdf

Pipeline reconstruction from fisheye images

A method of transforming systolic arrays using bit-parallel arithmetic into arrays using digit-serial arithmetic is described. Digit-serial computation is an area-time efficient method of doing high-speed arithmetic calculations, having the advantage through appropriate choice of digit and word size of allowing throughput capacity to be matched to design needs. For a certain class of systolic arrays, however, digit-serial arithmetic allows a further very significant benefit, by transforming arrays in which processors are under-utilized into arrays with 100% processor utilization. As an example, converting a well-known band-matrix multiplication array to use digit-serial processing is calculated to give an improvement of more than three times in area-time efficiency.

Systolic array processors for reducing under-utilization of original design parallel-bit processors with digit-serial processors by using maximum common divisor of latency around the loop connection

This paper presents a generalized version of the classic projective reconstruction theorem which helps to choose or assess depth constraints for projective depth estimation algorithms. The theorem shows that projective reconstruction is possible under a much weaker constraint than requiring all estimated projective depths to be nonzero. This result enables us to present classes of depth constraints under which any reconstruction of cameras and points projecting into given image points is projectively equivalent to the true camera-point configuration. It also completely specifies the possible wrong configurations allowed by other constraints. We demonstrate the application of the theorem by analysing several constraints used in the literature, as well as presenting new constraints with desirable properties. We mention some of the implications of our results on iterative depth estimation algorithms and projective reconstruction via rank minimization. Our theory is verified by running experiments on both synthetic and real data.

A Generalized Projective Reconstruction Theorem and Depth Constraints for Projective Factorization

We propose an object retrieval application which can retrieve user specified objects from a big supermarket. Significant and unpredictable scale difference between the query and the database image is the major obstacle encountered. The widely used local invariant features show their deficiency in such an occasion. To improve the situation, we first design a new weighting scheme which can assess the repeatability of local features against scale variance. Also, another method which deals with scale difference through retrieving a query under multiple scales is also developed. Our methods have been tested on a real image database collected from a local supermarket and outperform the existing local invariant feature based image retrieval approaches. A new spatial check method is also briefly discussed.

Richard Hartley

Papers

Bit-sliced digit-serial multiplier

Pipeline reconstruction from fisheye images

Systolic array processors for reducing under-utilization of original design parallel-bit processors with digit-serial processors by using maximum common divisor of latency around the loop connection

A Generalized Projective Reconstruction Theorem and Depth Constraints for Projective Factorization

Handling Significant Scale Difference for Object Retrieval in a Supermarket