Recognition of Banknotes in Multiple Perspectives Using Selective Feature Matching and Shape Analysis

Image-based localization plays an important role in today's autonomous driving technologies. However, in large scale outdoor environments, challenging conditions, e.g., lighting changes or different weather, heavily affect image appearance and quality. As a key component of feature-based visual localization, image feature detection and matching deteriorate severely and cause worse localization performance. In this paper, we propose a novel method for robust image feature matching under drastically changing outdoor environments. In contrast to existing approaches which try to learn robust feature descriptors, we train a deep network that outputs the low-rank representations of the images where the undesired variations on the images are removed, and perform feature extraction and matching on the learned low-rank space. We demonstrate that our learned low-rank images largely improve the performance of image feature matching under varying conditions over a long period of time.

In the field of image processing, a piece of information can be defined as a feature, which can be utilised to determine the computational task with references to some applications. Features may refer to specific structures in the images such as points, edges or objects. Feature detection and feature description are the initial steps of image registration process. There are many algorithms of feature detection like BRISK, FAST, SURF, etc. In our earlier paper [15], a hybrid algorithm for feature detection and feature description has been proposed. The next step of image registration is feature matching. This paper deals with feature matching algorithm using BRISK [4], FAST [3] and hybrid [15]. The proposed hybrid algorithm performs better results than the above two existing algorithms in terms of elapsed time, CPU time and performance measuring time. This can be clearly concluded that the feature matching result of hybrid in terms of tables and figures performs well. For testing the algorithm, river dam images of Odisha are chosen.

The algorithms based on image feature detection and matching are critical in the field of computer vision. Feature extraction and matching algorithms are used in many computer vision problems, including object recognition and structure from motion. Each feature detector and descriptor algorithm’s computational efficiency and robust performance have a major impact on image matching precision and time utilization. The performance of image matching algorithms that use expensive descriptors for detection and matching is addressed using existing approaches. The algorithm’s efficiency is measured by the number of matches found and the number of faults discovered when evaluated against a given pair of images. It also depends on the algorithm that detects the features and matches them in less amount of time. This paper examines and compares the different algorithm (SURF, ORB, BRISK, FAST, KAZE, MINEIGEN, MSER) performances using distinct parameters such as affine transformation, blur, scale, illumination, and rotation. The Oxford dataset is used to assess their robustness and efficiency against the parameters of interest. The time taken to detect features, the time taken to match images, the number of identified feature points, and the total running time is recorded in this study. The quantitative results show that the ORB and SURF algorithms detect and match more features than other algorithms. Furthermore, they are computationally less expensive and robust compared to other algorithms. In addition, the robustness of ORB and SURF is quite high in terms of outliers, and the amount of time taken to match with the reference is also significantly less. However, the efficiency of SURF reduces against blur transformation. FAST is good in detecting corners but lacks efficiency under different transformations. Experiments show that when each algorithm is subjected to numerous alterations, it has its own set of advantages and limitations.

In this paper we present a new shape analysis algorithm. The key distinguishing aspect of our algorithm is that it is completely compositional, bottom-up and non-iterative. We present our algorithm as an inference system for computing Hoare triples summarizing heap manipulating programs. Our inference rules are compositional: Hoare triples for a compound statement are computed from the Hoare triples of its component statements. These inference rules are used as the basis for a bottom-up shape analysis of programs.Specifically, we present a logic of iterated separation formula (LISF) which uses the iterated separating conjunct of Reynolds [17] to represent program states. A key ingredient of our inference rules is a strong bi-abduction operation between two logical formulas. We describe sound strong bi-abduction and satisfiability decision procedures for LISF.We have built a prototype tool that implements these inference rules and have evaluated it on standard shape analysis benchmark programs. Preliminary results show that our tool can generate expressive summaries, which are complete functional specifications in many cases.KeywordsModel CheckInference RuleAuxiliary VariableShape AnalysisCompose RuleThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Many computer vision applications rely on matching features of a query image to reference data sets, but little work has explored how quickly data sets become out of date. In this paper we measure feature matching performance across 5 years of time-lapse data from 20 static cameras to empirically study how feature matching is affected by changing sunlight direction, seasons, weather, and the structural changes over time in outdoor settings. We identify several trends that may be relevant in real world applications: (1) features are much more likely to match within a few days of the reference data, (2) weather and sun-direction have a large effect on feature matching, and (3) there is a slow decay over time due to physical changes in a scene, but this decay is much smaller than effects of lighting direction and weather. These trends are consistent across standard choices for feature detection (DoG, MSER) and feature description (SIFT, SURF, and DAISY). Across all choices, analysis of the feature detection and matching pipeline highlights that performance decay is mostly due to failures in key point detection rather than feature description.

Ever since the introduction of freely programmable hardware components into modern graphics hardware, graphics processing units (GPUs) have become increasingly popular for general purpose computations. Especially when applied to computer vision algorithms where a Single set of Instructions has to be executed on Multiple Data (SIMD), GPU-based algorithms can provide a major increase in processing speed compared to their CPU counterparts. This paper presents methods that take full advantage of modern graphics card hardware for real-time scale invariant feature detection and matching. The focus lies on the extraction of feature locations and the generation of feature descriptors from natural images. The generation of these featurevectors is based on the Speeded Up Robust Features (SURF) method [1] due to its high stability against rotation, scale and changes in lighting condition of the processed images. With the presented methods feature detection and matching can be performed at framerates exceeding 100 frames per second for 640 × 480 images. The remaining time can then be spent on fast matching against large feature databases on the GPU while the CPU can be used for other tasks. © 2008 IEEE.

Efficient detection and reliable matching of image features constitute a fundamental task in computer vision. When real-time operation is required, the solution to this problem becomes a real challenge, because of increased processing requirements. Scale Invariant Feature Transform (SIFT) is considered as a stable and robust algorithm for the extraction of invariant features, however special hardware is required for high frame-rate applications, in order to overcome its computational complexity in real-time. In this paper, a complete processor for SIFT feature matching in video sequences is presented. It performs SIFT feature extraction and matching, as well as rejection of false correspondences using the random sample consensus (RANSAC) algorithm. The processor was evaluated using metrics like response, repeatability and recall. It was found that the proposed fixed-point hardware implementation has a comparable performance with software implementations of SIFT. The processor architecture can process more than 36 fps with resolution 640×480, when it is implemented in Cyclone IV FPGA technology.

: This project is concerned with the study of advanced techniques for the analysis of reconnaissance imagery. The current project is concerned with the development and application of advanced techniques for image processing, feature detection, segmentation, texture and shape analysis, and region representation. (The overlays in the project title refer to region representations which specify background areas or possible object locations.)

We explore the performance of a number of popular feature detectors and descriptors in matching 3D object features across viewpoints and lighting conditions. To this end we design a method, based o...

Abstract. 3D reconstruction relies on accurate detection, extraction, description and matching of image features. This is even truer for complex architectural scenes that pose needs for 3D models of high quality, without any loss of detail in geometry or color. Illumination conditions influence the radiometric quality of images, as standard sensors cannot depict properly a wide range of intensities in the same scene. Indeed, overexposed or underexposed pixels cause irreplaceable information loss and degrade digital representation. Images taken under extreme lighting environments may be thus prohibitive for feature detection/extraction and consequently for matching and 3D reconstruction. High Dynamic Range (HDR) images could be helpful for these operators because they broaden the limits of illumination range that Standard or Low Dynamic Range (SDR/LDR) images can capture and increase in this way the amount of details contained in the image. Experimental results of this study prove this assumption as they examine state of the art feature detectors applied both on standard dynamic range and HDR images.

Many datasets have multiple perspectives – for example space, time and description – and often analysts are required to study these multiple perspectives concurrently. This concurrent analysis becomes difficult when data are grouped and split into small multiples for comparison. A design challenge is thus to provide representations that enable multiple perspectives, split into small multiples, to be viewed simultaneously in ways that neither clutter nor overload. We present a design framework that allows us to do this. We claim that multi‐perspective comparison across small multiples may be possible by superimposing perspectives on one another rather than juxtaposing those perspectives side‐by‐side. This approach defies conventional wisdom and likely results in visual and informational clutter. For this reason we propose designs at three levels of abstraction for each perspective. By flexibly varying the abstraction level, certain perspectives can be brought into, or out of, focus. We evaluate our framework through laboratory‐style user tests. We find that superimposing, rather than juxtaposing, perspective views has little effect on performance of a low‐level comparison task. We reflect on the user study and its design to further identify analysis situations for which our framework may be desirable. Although the user study findings were insufficiently discriminating, we believe our framework opens up a new design space for multi‐perspective visual analysis.

The computer vision technology is applied more and more extensively. In order to improve the effectiveness of various computer vision algorithms, the enhancement for low-illumination image is now playing an important role as image preprocessing. Aiming at the problems of image restoration and adaptive illumination adjustment, we proposed an improved low-illuminance image enhancement method based on convolutional neural network and Retinex theory. We first decompose the low-illuminance image into a reflection component and an illuminance component. Secondly, a convolutional neural network is designed for each of the two components. The reflection component branch restores the details of the image, and the illuminance component branch adaptively adjusts the image illuminance. Finally, the image is reconstructed. The experiments result show that our method has a better performance in SSIM and PSNR. Moreover, the image noise is well suppressed and the image details are also enhanced visually

Many computer vision applications rely on feature matching, hence the need for computationally efficient and robust 4D light field (LF) feature detectors and descriptors for applications using this imaging modality. In this paper, we propose a novel LF feature extraction method in the scale-disparity space, based on a Fourier disparity layer representation. The proposed feature extraction takes advantage of both the Harris feature detector and SIFT descriptor, and is shown to yield more accurate feature matching, compared with the LiFF light field feature with low computational complexity. In order to evaluate the feature matching performance with the proposed descriptor, we generated synthetic LF datasets with ground truth matching points. Experimental results with synthetic and real datasets show that, our solution outperforms existing methods in terms of both feature detection robustness and feature matching accuracy.

The detection and matching of point features play an important role in most of the computer vision algorithms, such as; for 3-D reconstruction and robotics navigation, localization and mapping. Over the last years various detectors and descriptors have been proposed and successfully applied to the different applications. However, the developed detectors are based on computationally intensive algorithms, so it is desirable to implement them in hardware platform based on high performance reconfigurable systems. Nowadays, due to the flexible structure of Field Programmable Gate Array (FPGA) device, the latter can be involved in such complex algorithms of image processing. In this paper, some related works concerning implementation of features detection and matching algorithms on FPGA will be discussed in the state of the art, then we will illustrate a general framework for implementing two common features extraction detectors and two features matching algorithms on FPGA-Nios II system.

Connected vehicles technology allows to share information between road elements, like vehicles, pedestrians and infrastructure. The main goal of the connected vehicles technology is to improve road safety and achieve autonomous driving. Many kinds of information can be shared over connected vehicles, like navigation information, road conditions and weather conditions. Videos and images are also important information to be shared over the connected vehicles network due to the rich information they carry to achieve autonomous driving, especially when they are processed by machine learning algorithms to detect and classify road objects. One of the issues in dealing with the images shared over connected vehicles is the image registration. Because the cameras used in vehicles have different specifications, like resolution, field of view and alignment. To analyze these images by computer vision algorithms, images have to be registered and aligned together. This paper proposes a solution to register the shared images that are collected by front vehicle cameras. Harris-Stephens approach for corner detection is used for feature extraction. The nearest neighbor distance ratio does the feature matching between the target image and the reference image. The algorithm calculates transformation parameters and applies them on target images to register them to a reference image. The paper evaluates the proposed system results using vehicle front camera collected dataset.