Robust Multiresolution Estimation of Parametric Motion Models

doi:10.1006/jvci.1995.1029

Journal of Visual Communication and Image Representation

Volume 6, Issue 4, December 1995, Pages 348-365

https://doi.org/10.1006/jvci.1995.1029 Get rights and content

Abstract

This paper describes a method to estimate parametric motion models. Motivations for the use of such models are, on one hand, their efficiency, which has been demonstrated in numerous contexts such as estimation, segmentation, tracking, and interpretation of motion, and on the other hand, their low computational cost compared to optical flow estimation. However, it is important to have the best accuracy for the estimated parameters, and to take into account the problem of multiple motion. We have therefore developed two robust estimators in a multi-resolution framework. Numerical results support this approach, as validated by the use of these algorithms on complex sequences.

References (0)

Cited by (446)

A level set based fractional order variational model for motion estimation in application oriented spectrum
2023, Expert Systems with Applications
In recent years, motion estimation has become prominent in the field of computer vision. This is due to its applications in various domains such as object detection, video surveillance, undersea navigation, fire damage control, particle image velocimetry etc. Therefore, in this paper, a nonlinear fractional order variational model is introduced for motion estimation in an image sequence (video). The motion estimation is performed in terms of optical flow. The objective of this work is to generalize the existing variational models from integer order to fractional order and provide the increased robustness against outliers, and furnish the dense and discontinuity preserving optical flow in various spectra such as synthetic, sintell, thermal, underwater, medical, fire and smoke and fluid image sequences. For this purpose, a level set segmentation based fractional order variational functional composed of a non-quadratic Charbonnier norm and a regularization term is propounded. This non-quadratic penalty provides an effective robustness against outliers, whereas the fractional derivative possesses a non-local character, and therefore is capable to deal with discontinuous information about texture and edges. The level set segmentation is performed on the flow field instead of images, which is a union of disjoint and independently moving regions such that each motion region contains objects of equal flow velocity. The numerical discretization of the fractional partial differential equations is employed with the help of Grünwald–Letnikov fractional derivative. The resulting nonlinear formulation is converted into a linear system and solved by an efficient numerical technique. The experimental results are carried out on 10 different application oriented spectra. The performance of the model is tested using different error measures and demonstrated against several outliers. The efficiency and accuracy of the proposed model is shown against recently published works. The graphical abstract for this algorithm is illustrated under the next section.
Video action recognition based on visual rhythm representation
2020, Journal of Visual Communication and Image Representation
Citation Excerpt :
Camera movement may introduce false motion and artifacts on video analysis. Jain et al. [10] make use of movement compensation [11] and optical flow to track a dense grid of points and adjust their positions so that camera movement is nullified. Experiments evaluate the impact of warping movement on each improved Dense Trajectory [8] descriptor.
Advances in video acquisition and storage technologies have promoted a great demand for automatic recognition of actions. The use of cameras for security and surveillance purposes has applications in several scenarios, such as airports, parks, banks, stations, roads, hospitals, supermarkets, industries, stadiums, schools. An inherent difficulty of the problem is the complexity of the scene under usual recording conditions, which may contain complex background and motion, multiple people on the scene, interactions with other actors or objects, and camera motion. Most recent databases are built primarily with shared recordings on YouTube and with snippets of movies, situations where these obstacles are not restricted. Another difficulty is the impact of the temporal dimension since it expands the size of the data, increasing computational cost and storage space. In this work, we present a methodology of volume description using the Visual Rhythm (VR) representation. This technique reshapes the original volume of the video into an image, where two-dimensional descriptors are computed. We investigated different strategies for constructing the representation by combining configurations in several image domains and traversing directions of the video frames. From this, we propose two feature extraction methods, Naïve Visual Rhythm (Naïve VR) and Visual Rhythm Trajectory Descriptor (VRTD). The first approach is the straightforward application of the technique in the original video volume, forming a holistic descriptor that considers action events as patterns and formats in the visual rhythm image. The second variation focuses on the analysis of small neighborhoods obtained from the process of dense trajectories, which allows the algorithm to capture details unnoticed by the global description. We tested our methods in eight public databases, one of hand gestures (SKIG), two in first person (DogCentric and JPL), and five in third person (Weizmann, KTH, MuHAVi, UCF11 and HMDB51). The results show that the developed techniques are able to extract motion elements along with format and appearance information, achieving competitive accuracy rates compared to state-of-the-art action recognition approaches.
Novel remapping approach for HR-EBSD based on demons registration
2020, Ultramicroscopy
In this study, the possibility of utilizing a computer vision algorithm, i.e., demons registration, to accurately remap electron backscatter diffraction patterns for high resolution electron backscatter diffraction (HR-EBSD) applications is presented. First, the angular resolution of demons registration is demonstrated to be lower than the conventional cross-correlation based method, particularly at misorientation angles >0.157 rad. In addition, GPU acceleration has been applied to significantly boost the speed of iterative registration between a pair of patterns with 0.175 rad misorientation to under 1 s. Second, demons registration is implemented as a first-pass remapping, followed by a second pass cross-correlation method, which results in angular resolution of ~0.5 × 10⁻⁴ rad, a phantom stress value of ~35 MPa and phantom strain of ~2 × 10⁻⁴, on dynamically simulated patterns, without the need of implementing robust fitting or iterative remapping. Lastly, the new remapping method is applied to a large experimental dataset collected from an as-built additively-manufactured Inconel 625 cube, which shows significant residual stresses built-up near the large columnar grain region and regularly arranged GND structures.
CNN-based temporal detection of motion saliency in videos
2019, Pattern Recognition Letters
The problem addressed in this paper appertains to the domain of motion saliency in videos. However this is a new problem since we aim to extract the temporal segments of the video where motion saliency is present. It turns out to be a frame-based classification problem. A frame will be classified as dynamically salient if it contains local motion departing from its context. Temporal motion saliency detection is relevant for applications where one needs to trigger alerts or to monitor dynamic behaviours from videos. It can also be viewed as a prerequisite before computing motion saliency maps. The proposed approach handles situations with a mobile camera. It involves two main stages consisting first in cancelling the global motion due to the camera movement, then in applying a deep learning classification framework. We have investigated two ways of implementing the first stage, based on image warping, and on residual flow respectively. Experiments on real videos demonstrate that we can obtain accurate classification in highly challenging situations.
A survey of variational and CNN-based optical flow techniques
2019, Signal Processing: Image Communication
Dense motion estimations obtained from optical flow techniques play a significant role in many image processing and computer vision tasks. Remarkable progress has been made in both theory and its application in practice. In this paper, we provide a systematic review of recent optical flow techniques with a focus on the variational method and approaches based on Convolutional Neural Networks (CNNs). These two categories have led to state-of-the-art performance. We discuss recent modifications and extensions of the original model, and highlight remaining challenges. For the first time, we provide an overview of recent CNN-based optical flow methods and discuss their potential and current limitations.
Automatic 3D-display-friendly scene extraction from video sequences and optimal focusing distance identification
2024, Multimedia Tools and Applications

View all citing articles on Scopus

View full text

Regular ArticleRobust Multiresolution Estimation of Parametric Motion Models

Abstract

Regular Article
Robust Multiresolution Estimation of Parametric Motion Models