SURF Features Research Articles

An efficient ear recognition technique invariant to illumination and pose

This paper presents an efficient ear recognition technique which derives benefits from the local features of the ear and attempt to handle the problems due to pose, poor contrast, change in illumination and lack of registration. It uses (1) three image enhancement techniques in parallel to neutralize the effect of poor contrast, noise and illumination, (2) a local feature extraction technique (SURF) on enhanced images to minimize the effect of pose variations and poor image registration. SURF feature extraction is carried out on enhanced images to obtain three sets of local features, one for each enhanced image. Three nearest neighbor classifiers are trained on these three sets of features. Matching scores generated by all three classifiers are fused for final decision. The technique has been evaluated on two public databases, namely IIT Kanpur ear database and University of Notre Dame ear database (Collections E). Experimental results confirm that the use of proposed fusion significantly improves the recognition accuracy.

Automated sparse 3D point cloud generation of infrastructure using its distinctive visual features

The commercial far-range (>10 m) spatial data collection methods for acquiring infrastructure’s geometric data are not completely automated because of the necessary manual pre- and/or post-processing work. The required amount of human intervention and, in some cases, the high equipment costs associated with these methods impede their adoption by the majority of infrastructure mapping activities. This paper presents an automated stereo vision-based method, as an alternative and inexpensive solution, to producing a sparse Euclidean 3D point cloud of an infrastructure scene utilizing two video streams captured by a set of two calibrated cameras. In this process SURF features are automatically detected and matched between each pair of stereo video frames. 3D coordinates of the matched feature points are then calculated via triangulation. The detected SURF features in two successive video frames are automatically matched and the RANSAC algorithm is used to discard mismatches. The quaternion motion estimation method is then used along with bundle adjustment optimization to register successive point clouds. The method was tested on a database of infrastructure stereo video streams. The validity and statistical significance of the results were evaluated by comparing the spatial distance of randomly selected feature points with their corresponding tape measurements.

Dense statistic versus sparse feature-based approach for 3D object recognition

In this article we introduce and compare two approaches towards automatic classification of 3D objects in 2D images. The first one is based on statistical modeling of wavelet features. It estimates probability density functions for all possible object classes considered in a particular recognition task. The second one uses sparse local features. For training, SURF features are extracted from the training images. During the recognition phase, features from the image are matched geometrically, providing the best fitting object for the query image. Experiments were performed for different training sets using more than 40 000 images with different backgrounds. Results show very good classification rates for both systems and point out special characteristics for each approach, which make them more suitable for different applications.

Human Pose Tracking For Video Using SURF features

In this paper, a novel method based on SURF features method for tracking human motion in monocular videos is proposed. With a initial human skeleton joint point template, we use the probability density propagation of the particle filers through the model. This algorithm can automatically achieve right human motion figure from tracking failures, such as occlusion and auto-occlusion problem. Experimental results from 20 classes monocular videos show that the new Based on SURF method is robust and the tracking results are good.

크기 및 회전 불변 특징점을 이용한 파노라마 영상 합성 알고리즘

본 논문은 동일한 대상물을 촬영한 영상을 합성하여 파노라마 영상을 생성하는 방법에 대하여 설명한다. 디지털 카메라의 보급으로 파노라마 영상에 대한 관심이 높아지면서 다양한 방법의 파노라마 영상의 제작 방법이 연구되고 있다. 본 논문에서는 크기 및 회전 불변 특징점을 활용하여 파노라마 영상을 합성하는 방법에 대해서 제안한다. 먼저, 입력 영상들에 대해서 특징점을 추출하고, RANSAC 알고리즘을 통해 추출된 특징점을 정합한다. 정합점을 이용하여 투영 변환식을 모델링하고, 모델링된 변환식을 통하여 영상을 정렬하여 파노라마 영상을 생성한다. 제안한 알고리즘은 SURF 특징점 추출 알고리즘을 적용하여 영상의 크기 및 회전 등의 기하학적 변형에 강인하며, 처리 속도도 향상하였다. 실험에서는 기존 Harris corner 검출기나 SIFT 알고리즘을 통해 검출한 특징과 제안한 알고리즘에서 사용된 SURF 알고리즘을 비교 분석 하였고, <TEX>$640{\times}480$</TEX> 크기의 영상을 이용하여 제안한 알고리즘을 통해 파노라마 영상을 합성하였다. 그 결과 파노라마 영상의 합성에 소요되는 시간은 평균0.4초로 나타났고, 기존 알고리즘에 비하여 효율적인 것으로 나타났다. This paper addresses the way to compose paronamic images from images taken the same objects. With the spread of digital camera, the panoramic image has been studied to generate with its interest. In this paper, we propose a panoramic image generation method using scaling and rotation invariant features. First, feature points are extracted from input images and matched with a RANSAC algorithm. Then, after the perspective model is estimated, the input image is registered with this model. Since the SURF feature extraction algorithm is adapted, the proposed method is robust against geometric distortions such as scaling and rotation. Also, the improvement of computational cost is achieved. In the experiment, the SURF feature in the proposed method is compared with features from Harris corner detector or the SIFT algorithm. The proposed method is tested by generating panoramic images using <TEX>$640{\times}480$</TEX> images. Results show that it takes 0.4 second in average for computation and is more efficient than other schemes.

A Method for Counting Moving People in Video Surveillance Videos

People counting is an important problem in video surveillance applications. This problem has been faced either by trying to detect people in the scene and then counting them or by establishing a mapping between some scene feature and the number of people (avoiding the complex detection problem). This paper presents a novel method, following this second approach, that is based on the use of SURF features and of an -SVR regressor provide an estimate of this count. The algorithm takes specifically into account problems due to partial occlusions and to perspective. In the experimental evaluation, the proposed method has been compared with the algorithm by Albiol et al., winner of the PETS 2009 contest on people counting, using the same PETS 2009 database. The provided results confirm that the proposed method yields an improved accuracy, while retaining the robustness of Albiol's algorithm.

Analysis of Feature Point Distributions for Fast Image Mosaicking Algorithms

In many algorithms the registration of image pairs is done by feature point matching. After the feature detection is performed, all extracted interest points are usually used for the registration process without further feature point distribution analysis. However, in the case of small and sparse sets of feature points of fixed size, suitable for real-time image mosaicking algorithms, a uniform spatial feature distribution across the image becomes relevant. Thus, in this paper we discuss and analyze algorithms which provide different spatial point distributions from a given set of SURF features. The evaluations show that a more uniform spatial distribution of the point matches results in lower image registration errors, and is thus more beneficial for fast image mosaicking algorithms.

FASTSLAM USING SURF FEATURES: AN EFFICIENT IMPLEMENTATION AND PRACTICAL EXPERIENCES

This paper describes how the recently published SURF features can be used as landmarks for an online FastSLAM algorithm that simultaneously estimates the robot pose and the pose of a large number of landmarks. An implementation with particular focus on efficient data structures like two-stage landmark data base and special balanced binary trees is described. Practical results on outdoor data sets at 3 Hz with about 3-6 % error of total traveled distance are shown.

Vertical profiles and horizontal scales of bubble clouds in the surf zone as measured with a distributed array of upward-looking sonars

A collaborative, multi-institute experiment was recently performed in the vicinity of the Scripps Pier, whose goal was the study of bubble-field generation, transport, and distribution, as influenced by surf zone conditions. The collaboration produced an ensemble of instrumentation at the site for measuring bubbles, ambient noise, temperature and salinity, currents, and surface waves. This paper emphasizes data gathered by APL-UW using a set of four upward-looking transducers (frequency 240 kHz), which simultaneously measured vertical profiles of acoustic volume scattering from bubbles from four locations. The primary locus of measurement activity was defined by NRL-SSC’s triangular frame (Delta Frame) and the four transducer locations for the APL-UW system bracketed this region. The transport of bubbles via rip currents emerged as a key feature surf zone bubble environment. Images of volumetric backscattering strength versus time and depth reveal episodic events (by way of increased scattering level) which last O(100) s and can occur in intervals of O(10) min. Time lags for increased scattering at the four locations were consistent with a seaward flow of rip currents equal to ∼0.25 m/s. Other features of the data that complement the single-depth measurements made from the Delta Frame will be discussed. [Work supported by ONR.]