Extract Interest Points Research Articles

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.

1P1-C27 自律型水中ロボットにおける自動モザイク画像生成システム

Underwater robots are expected as the attractive tool for ocean development and investigation. For land robot, image processing is one of the most efficient data extraction methods. However, for underwater robots, image processing is not so often utilized because of darkness of underwater environment, unclearness of vision, and payload limitation for image processing devices. In this paper, we developed an automatic underwater video mosaic system. At first, we extract interest points from image obtained by underwater vehicle. Second, this system tracked extracted features over adjacent images. Finally, we calculate homography matrix to stitch images. We aim to develop a robust mosaicing system for dynamic environment such as swimming fish and bubbles.

Simulating gaze attention behaviors for crowds

AbstractCrowd animation is a topic of high interest which offers many challenges. One of the most important is the trade‐off between rich, realistic behaviors, and computational costs. To this end, much effort has been put into creating variety in character representation and animation. Nevertheless, one aspect still lacking realism in virtual crowd characters resides in their attention behaviors. In this paper, we propose a framework to add gaze attention behaviors to crowd animations. First, we automatically extract interest points from character or object trajectories in pre‐ existing animations. For a given character, we assign a set of elementary scores based on parameters such as distance or speed to all other characters or objects in the scene. We then combine these subscores in an overall scoring function. The scores obtained from this function form a set of gaze constraints that determine where and when each character should look. We finally enforce these constraints with an optimized dedicated gaze Inverse Kinematics (IK) solver. It first computes the displacement maps for the constraints to be satisfied. It then smoothly propagates these displacements over an automatically defined number of frames. We demonstrate the efficiency of our method and our visually convincing results through various examples. Copyright © 2009 John Wiley & Sons, Ltd.

TH‐C‐AUD C‐02: Automatic Extraction of Salient Interest Points in 3D Images for Contour Propagation in IGRT

Purpose: To efficiently propagate patient contours to multi‐fraction Cone‐Beam CT volumes. Method and Materials: An auto‐segmentation scheme was developed which uses anchor points to propagate structure delineations from one image to another, accounting for the changing patient anatomy. Four patients involved in 35‐ or 40‐fraction Head‐and‐Neck treatment schemes were selected retrospectively. The 8–12 Organ‐at‐Risk contours drawn in their initial planning CT included spinal cord, brainstem, optic structures, and parotids, to be propagated to the CBCT image acquired at each fraction. Our method involves three steps: a) Generate a patient‐specific compressed model in terms of salient points from the planning CT image. The automatic extraction of these uses a developed 3‐dimensional extension of the SIFT algorithm; b) Retrieve this model in every CBCT dataset, via automatic block‐matching maximizing the local correlation between neighborhoods around the interest points; c) Propagate the original contours based on the thin‐plate‐spline warping transformation derived from salient point pairs. Results: For each patient, over 500 salient points were extracted in 5–7 minutes, and subsequently retrieved within 45 seconds in every CBCT image, into which the 8–12 contours were propagated in 10 seconds. The results show clear improvement, compared to contours obtained by both a mere copy and voxel‐based rigid registration; the typical neck shrinking and spine flexions especially are followed. Our method is more robust and computationally affordable than voxel‐based deformable image registration. Conclusion: The contours can be successfully propagated to all fractions based on significant salient anchor points extracted to summarize one patient's gray‐level information and geometry. The thus highlighted deformations of the critical structures i) can help determine a safe dose map and be reflected in Planning organ at Risk Volume (PRV) margins, ii) may be used to support time‐efficient re‐planning in adaptive radiotherapy. Conflict of Interest: This research is partially supported by Philips Healthcare.

Error evaluation of scale‐invariant local descriptor and its application to image indexing

AbstractOne technique for image indexing is the method of extracting interest points from an image and finding point‐to‐point matches with descriptor vectors which characterize the vicinity of those interest points. In this paper, in order to improve the accuracy of these point‐to‐point matches, the authors introduce a new distance measure to replace the Mahalanobis distance used before now as the measure for the distance between descriptor vectors. The number of mismatches which tend to result from establishing matches using the Mahalanobis distance can be reduced by finding the observational error distribution of the descriptors using synthetic images including errors and revising the distance measure based thereon. Also, among the attributes of interest points, the relatively high reproducibility of the characteristic scale is of note. A technique is proposed which makes active use of characteristic scale as a descriptor, as well as Local Jet which was used before now, and the effectiveness thereof for indexing is demonstrated. © 2006 Wiley Periodicals, Inc. Electron Comm Jpn Pt 3, 90(2): 31–39, 2007; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecjc.20310

Affine invariant pattern recognition using multiscale autoconvolution

This paper presents a new affine invariant image transform called Multiscale Autoconvolution (MSA). The proposed transform is based on a probabilistic interpretation of the image function. The method is directly applicable to isolated objects and does not require extraction of boundaries or interest points, and the computational load is significantly reduced using the Fast Fourier Transform. The transform values can be used as descriptors for affine invariant pattern classification and, in this article, we illustrate their performance in various object classification tasks. As shown by a comparison with other affine invariant techniques, the new method appears to be suitable for problems where image distortions can be approximated with affine transformations.

点群データからの特徴点の効率的抽出と歴史的建造物の三次元モデリング手法に関する研究

Recently, laser scanners have been receiving more attention as a useful tool for real-time 3D data acquisition, and various applications such as city modeling, DTM generation and 3D modeling of cultural heritage were proposed. However, 3D modeling using point cloud data are still issues. In particular, efficient extraction of interest points for generating TIN model should be resolved.With this motive, efficient extraction of interest points in point crowd data and 3D modeling of historical structure are investigated in this paper. The method is based on surface flatness, and surface is classified into flat area (ground surface, wall of buildings, etc.) and non flat area (trees, bushes etc.) . Furthermore, break-line is detected as boundary between flat and non flat area, and each flat area is segmented using break-lines. The most remarkable point of this method is its ability to extract interest points for generating TIN model using 3D information for each point on the detected break line.

Image matching with scale adjustment

In this paper we address the problem of matching two images with two different resolutions: a high-resolution image and a low-resolution one. The difference in resolution between the two images is not known and without loss of generality one of the images is assumed to be the high-resolution one. On the premise that changes in resolution act as a smoothing equivalent to changes in scale, a scale-space representation of the high-resolution image is produced. Hence the one-to-one classical image matching paradigm becomes one-to-many because the low-resolution image is compared with all the scale-space representations of the high-resolution one. Key to the success of such a process is the proper representation of the features to be matched in scale-space. We show how to represent and extract interest points at variable scales and we devise a method allowing the comparison of two images at two different resolutions. The method comprises the use of photometric- and rotation-invariant descriptors, a geometric model mapping the high-resolution image onto a low-resolution image region, and an image matching strategy based on local constraints and on the robust estimation of this geometric model. Extensive experiments show that our matching method can be used for scale changes up to a factor of 6.

Context-free attentional operators: The generalized symmetry transform

Active vision systems, and especially foveated vision systems, depend on efficient attentional mechanisms. We propose that machine visual attention should consist of both high-level, context-dependent components, and low-level, context free components. As a basis for the context-free component, we present an attention operator based on the intuitive notion of symmetry, which generalized many of the existing methods of detecting regions of interest. It is a low-level operator that can be applied successfully without a priori knowledge of the world. The resultingsymmetry edge map can be applied in various low, intermediate-and high- level tasks, such as extraction of interest points, grouping, and object recognition. In particular, we have implemented an algorithm that locates interest points in real time, and can be incorporated in active and purposive vision systems. The results agree with some psychophysical findings concerning symmetry as well as evidence concerning selection of fixation points. We demonstrate the performance of the transform on natural, cluttered images.