A high performance parallel computing architecture for robust image features

Speeded up robust features (SURF) is a state of the art computer vision algorithm that relies on integral image representation for performing fast detection and description of image features that are scale and rotation invariant. Integral image representation, however, has major draw back of large binary word length that leads to substantial increase in memory size. When designing a dedicated hardware to achieve real-time performance for the SURF algorithm, it is imperative to consider the adverse effects of integral image on memory size, bus width and computational resources. With the objective of minimizing hardware resources, this paper presents a novel implementation concept of a reduced word length integral image based SURF detector. It evaluates two existing word length reduction techniques for the particular case of SURF detector and extends one of these to achieve more reduction in word length. This paper also introduces a novel method to achieve integral image word length reduction for SURF detector.

Speeded Up Robust Features (SURF) algorithm is a fast robust local feature matching method. Because of the advantage of invariance of scale and rotation, the SURF algorithm has been widely used in the image registration. However, due to the acquired angle, imaging mode, different resolution, SURF algorithm becomes quite difficult in the image registration. This paper presents a normalized SURF algorithm which can reduce the impact of hue difference between remote sensing images. Combined with the idea of two-way matching, the normalized SURF is able to improve the accuracy of the image registration. The experimental results show the normalized SURF algorithm can keep greater accuracy and relatively higher speed of matching than the original SURF algorithm.

위성 SAR 영상의 활용이 증가하면서 영상의 해석 및 융합을 위한 정밀 기하보정에 대한 필요성이 높아지고 있다. 특히 광역감시 목적으로 활용되기 위해 서로 다른 해상도를 갖는 SAR 영상간 정보융합도 활발해지고 있다. 일반적으로 SAR 영상의 기하보정은 위성의 궤도 및 자세정보를 활용하여 수행할 수 있지만 SAR 센서의 궤도 및 시스템 오차, 대상지형 특성에 의한 왜곡으로 인해 추가적인 보정이 필요하게 된다. SAR 영상을 통한 변화탐지나 타 영상과의 융합에 적용하기 위해서는 기하 오차 보정이 반드시 선행되어야 한다. 이를 위해 다수의 지상 기준점을 선정하고 이를 포함하는 기준 영상과 비교하여 원본 영상에서 대응점을 찾는 방식으로 정밀 기하보정을 수행할 수 있다. Speeded Up Robust Feature (SURF) 기법은 쉽고 빠르게 영상의 기준점을 찾을 수 있지만 상대적으로 해상도가 낮고 스펙클 잡음에 영향을 받는 SAR 영상에서는 활용하기가 어렵다고 알려져 있다. 본 논문에서는 SURF 기법을 위성 SAR 영상에 적용할 때 발생할 수 있는 오차를 추출하고 영상 특성에 따른 성능 변화를 분석하였다. SURF 알고리즘의 적용이 가능한 입력 변수의 적정 범위를 제시하고 그에 따른 영상 정합의 오차를 분석하여 중저해상도의 위성 SAR 영상에 대해서도 SURF 기법을 통한 기하 보정 및 영상 정합이 적용될 수 있음을 검증하였다. As applications of spaceborne SAR imagery are extended, there are increased demands for accurate registrations for better understanding and fusion of radar images. It becomes common to adopt multi-resolution SAR images to apply for wide area reconnaissance. Geometric correction of the SAR images can be performed by using satellite orbit and attitude information. However, the inherent errors of the SAR sensor's attitude and ground geographical data tend to cause geometric errors in the produced SAR image. These errors should be corrected when the SAR images are applied for multi-temporal analysis, change detection applications and image fusion with other sensor images. The undesirable ground registration errors can be corrected with respect to the true ground control points in order to produce complete SAR products. Speeded Up Robust Feature (SURF) technique is an efficient algorithm to extract ground control points from images but is considered to be inappropriate to apply to SAR images due to high speckle noises. In this paper, an attempt is made to apply SURF algorithm to SAR images for image registration and fusion. Matched points are extracted with respect to the varying parameters of Hessian and SURF matching thresholds, and the performance is analyzed by measuring the imaging matching accuracies. A number of performance measures concerning image registration are suggested to validate the use of SURF for spaceborne SAR images. Various simulations methodologies are suggested the validate the use of SURF for the geometric correction and image registrations and it is shown that a good choice of input parameters to the SURF algorithm should be made to apply for the spaceborne SAR images of moderate resolutions.

Copy-Move forgery is the most widely used way to tamper image for creating the forged images. The meaning of the photograph is changed by forging or concealing the images. Copy move is a very common technique for image forgery; tools used for retouching are applied along with it making naked eyes difficult to detect the copy move. Hence, the verification of authenticity and integrity is very essential of the image. The copy-move forgery detection can be divided into two heads a) Block based method b) Keypoint based. In the case of blocked based method, the method is exceptionally good in detecting the forged regions with higher accuracy but uses extremely high power of computation. Speed up Robust Feature (SURF) and Scale Invariant Feature Transform (SIFT) are invariant with respect to geometrical and illumination transform. In this paper we review the different keypoint approach which is a better approach or alternate to block-based approach due to invariancy property of SURF and SIFT. The keypoint based copy-move forgery detection schemes involve, identifying and relating the localized features of the images using the algorithms like SIFT and SURF.

This paper presents an efficient feature detection algorithm based on the classical SURF (Speeded Up Robust Feature) detector. The image features are represented and scored with respect to its local symmetry property. The local symmetry has natural properties of scale and transformation invariants, and also insensitive to illumination change and local noise. By the proposed feature descriptor, the calculation of 64-dimensional vectors in SURF algorithm can be reduced to 16-dimensional vector respectively. The local symmetry score is defined as the sum of minimum distance between each feature point and its neighboring points in an image based on the image intensities. The algorithm is experimented with some real images and the results are compared with the original SURF algorithm to show its improvement.

영상에서 임의의 점에 대한 고유한 특징을 계산하는 알고리즘은 파노라마 영상의 제작, 스테레오 영상의 획득, 물체 인식, 이미지 분석 등에 다양하게 사용되는 중요한 요소이다. 일반적으로 어떤 점의 특징은 스칼라 형태가 아닌 벡터형태로 나타나게 되는데, 무수히 많은 특징 점들을 서로 비교하는 작업은 매우 많은 계산량을 요구한다. 본 연구에서는 영상의 특징점 계산에 SURF(speeded up robust features)를 이용하였고, 이미지로부터 추출된 특징을 PCA(principal component analysis)기법을 이용하여 벡터의 차원을 축소하여 연결리스트 자료구조에 정렬한 다음 특징을 비교하는 기법을 제안한다. 제안된 특징의 비교 방법을 적용할 경우 기존 방법의 매칭 정확도는 유지한 상태에서 계산시간을 줄일 수 있는 것을 실험을 통하여 확인하였다. Extracting unique features from an image is a fundamental issue when making panorama images, acquiring stereo images, recognizing objects and analyzing images. Generally, the task to compare features to other images requires much computing time because some features are formed as a vector which has many elements. In this paper, we present a method that compares features after reducing the feature dimension extracted from an image using PCA(principal component analysis) and sorting the features in a linked list. SURF(speeded up robust features) is used to describe image features. When the dimension reduction method is applied, we can reduce the computing time without decreasing the matching accuracy. The proposed method is proved to be fast and robust in experiments.

How to construct compact and powerful feature descriptors is an important research subject in the fields of machine vision and pattern recognition. After the research on SURF, the issue that the Haar descriptor of Speeded Up Robust Features(SURF) algorithm can not make full use of the information around the neighborhood of the feature points is found. To resolve the issue, this paper proposes a improved SURF based on the original SURF algorithm. The algorithm makes full use of the information around the neighborhood of the sub-block by masking with 3*3 window templates and then constructs the descriptor which can obtain better discriminative power. In addition, aimed at the primary direction of the SURF depends too much on the local region, the paper improves the confirmation of the primary direction. At last, this paper combines the nearest distance and the ratio between the nearest distance and the second nearest distance to match the interest points. The experiment shows that the algorithm proposed in this paper has better discriminative power, and the face recognition rate is enhanced.

— Promotion is an attempt to notify or offer a product or service with the aim of attracting potential customers to buy or consume it, with the promotion, producers or distributors expect an increase in sales figures. In this study researchers used Augmented Reality technology for interactive media promotion of food menus by adding 3D multimedia elements. The method used in this study uses the Natural Feature Tracking method with the Speed Up Robust Features (SURF) algorithm, which detects local features in marker images that are resistant to rotation, scale and blurring. The results showed that the keypoint functions to render 3D objects. Search for keypoints is interrupted due to distance, light intensity and slope of the marker. Test results to see the distance between the camera and the marker as far as 60 cm. Medium light intensity that can detect markers, the average time of object speed can be displayed is 3.026 seconds and the marker slope limit is 30 °. This is because the keypoint readings at the position and time limit from keypoint readings clearly, keypoint readings clearly produce 3D objects can be displayed. This research uses the Android platform as the foundation of this Augmented Reality technology application. So that by displaying 3D food menu items in restaurants it is expected to be a means of promotion to attract consumers.

Feature description is a critical task in Augmented Reality Tracking. This article introduces a Convex Based Feature Descriptor (CBFD) system designed to withstand rotation, lighting, and blur variations while remaining computationally efficient. We have developed two filters capable of computing pixel intensity variations, followed by the covariance matrix of the polynomial to describe the features. The superiority of CBFD is validated through precision, recall, computation time, and feature location distance. Additionally, we provide a solution to determine the optimal block size for describing nonlinear regions, thereby enhancing resolution. The results demonstrate that CBFD achieves a average precision of 0.97 for the test image, outperforming Superpoint, Directional Intensified Tertiary Filtering (DITF), Binary Robust Independent Elementary Features (BRIEF), Binary Robust Invariant Scalable Keypoints (BRISK), Speeded Up Robust Features (SURF), and Scale Invariant Feature Transform (SIFT), which achieve scores of 0.95, 0.92, 0.72, 0.66, 0.63 and 0.50 respectively. Noteworthy is CBFD's recall value of 0.87 representing at the maximum of a 13.6% improvement over Superpoint, DITF, BRIEF, BRISK, SURF, and SIFT. Furthermore, the matching score for the test image is 0.975. The computation time for CBFD is 2.8 ms, which is at least 6.7% lower than that of other algorithms. Finally, the plot of location feature distance illustrates that CBFD exhibits minimal distance compared to DITF and Histogram of Oriented Gradients (HOG). These results highlight the speed and robustness of CBFD across various transformations.

This paper proposes a feature-based image registering technique for remote sensing satellite images. The system is composed of multiple steps in which various image processing techniques are used. Speed-up Robust Features (SURF) and Scale Invariant Feature Transform (SIFT) can be mounted alternatively to detect and define features on images. We also address the issues of registering multi-spectral and multi-temporal images. The effectiveness of the proposed technique has been studied by registering partially overlapping satellite mosaic images of the Marmara Sea. The images were captured by the recently-launched LandSat-8 satellite.

In recent years, keypoints-based features, such as speeded up robust features (SURF), capture the attention of researchers in the area of image copy-move forgery detection. They are fast to compute and are quite robust to most inter-processing and post-processing operations. However, they will fail to detect the forgery if not enough keypoints are detected, and that is the case when the forgery size is small. A small size forgery is possible if the object is far away from the camera, or if the image is scaled down as a post-processing step. In this paper, our focus is on detecting forgeries of small size using an improved SURF based copy-move forgery detection (CMFD) method. Our method improves keypoints detection by preprocessing the image using a single image super resolution (SISR) algorithm. The proposed approach has been evaluated and compared against original SURF through a comprehensive set of experiments using a dataset of small size forgeries. Experimental results indicate that our method outperforms SURF especially when the forgery size is small.

In order to improve the speed of stitching and information extraction of remote sensing images, based on the UAV (Unmanned Aerial Vehicle) image stitching algorithm, this paper proposes a UAV image matching method based on the combination of SURF (Speeded Up Robust Features) and MASC (M-Estimate Sample Consensus) algorithm. the SURF algorithm is fast, and the MASC algorithm can effectively eliminate mismatch points and has high stability. In this paper, the SIFT (Scale-invariant feature transform) algorithm, SURF algorithm, SURF and RANSAC (Random Sample Consensus) algorithm as well as SURF and MASC algorithm are used to stitch the same two images and compare them respectively. The experimental results show that the lowest image matching error rate is only 2.13% and the matching time is 4.772670s when using the combination of SURF and MASC algorithm for image stitching, which has certain advantages.

Image registration is a vital problem in medical imaging. It has many potential applications in clinical diagnosis (e.g. diagnosis of cardiac, retinal, pelvic, renal, abdomen, liver and tissue disorders). In this paper a new technique has been proposed based on feature based image registration using SURF (Speeded up robust features) and Hessian matrix feature extraction techniques for detecting, extracting and matching corresponding points in the reference (fixed) and source image (deformed). The source image is aligned using SURF and Hessian feature extraction techniques to align with the reference image. The proposed technique is based on multimodality of images, feature based classification and non-rigid images as the images taken are from different sources like MRI, CT scan and X-ray. The proposed technique uses feature-based methods to find correspondence between image features such as points, lines and contours. The non-rigid images use fixed (source) image and also deformable (target) image of the subject and aligns target image according to source/reference image.

The recently proposed, KAZE image feature detection and description algorithm (Alcantarilla et al. in Proceedings of the British machine vision conference. LNCS, vol 7577, no 6, pp 13.1–13.11, 2013) offers significantly improved robustness in comparison to conventional algorithms like SIFT (scale-invariant feature transform) and SURF (speeded-up robust features). The improved robustness comes at a significant computational cost, however, limiting its use for many applications. We report a GPU acceleration of the KAZE algorithm that is significantly faster than its CPU counterpart. Unlike previous reports, our acceleration does not resort to binary descriptors and can serve as a drop-in replacement for CPU-KAZE, SIFT, SURF etc. By achieving nearly tenfold speedup (for a 1920 by 1200 sized image, our Compute Unified Device Architecture (CUDA)-C implementation took around 245 ms on a single GPU in comparison to nearly 2400 ms for a 16-threaded CPU version) without degradation in feature extraction performance, our work expands the applicability of the KAZE algorithm. Additionally, the strategies described here could also prove useful for the GPU implementation of other nonlinear scale-space-based image processing algorithms.

The emergence of social media has provided vast amounts of information that is potentially valuable for emergency management. In the EU-FP7 Project Security Systems for Language and Image Analysis (Slandail), an image analysis system has been developed to recognize the flood water images from the social media resources by incorporating with text analysis. A novel image feature descriptor has been developed to facilitate fast image processing based on incorporation of the "Squiral" (Square-Spiral) Image Processing (SIP) framework with the "Speeded-up Robust Features" (SURF). A new approach is proposed to generate an index from image recognition outcomes based on a moving window average, which presents a temporal change based on the occurrence of flooding water identified by image analysis. The evaluation for computation time and recognition were based on a batch of images obtained from the US Federal Emergency Management Agency (FEMA) media library and Facebook corpus from Germany, and the outcomes show the advantages of the proposed image features. The simulation results demonstrate the concept of the index based on a moving window average, highlighting the potential for application in emergency management.