Active Rays: Polar-transformed Active Contours for Real-Time Contour Tracking

doi:10.1006/rtim.1997.0116

Real-Time Imaging

Volume 5, Issue 3, June 1999, Pages 203-213

https://doi.org/10.1006/rtim.1997.0116 Get rights and content

Abstract

In this paper we describe a new approach to contour extraction and tracking, which is based on the principles of active contour models and overcomes its shortcomings. We formally introduce active rays, describe the contour extraction as an energy minimization problem and discuss what active contours and active rays have in common.

The main difference is that for active rays a unique ordering of the contour elements in the 2D image plane is given, which cannot be found for active contours. This is advantageous for predicting the contour elements' position and prevents crossings in the contour. Further, another advantage is that instead of an energy minimization in the 2D image plane the minimization is reduced to a 1D search problem. The approach also shows any-time behavior, which is important with respect to real-time applications. Finally, the method allows for the management of multiple hypotheses of the object's boundary. This is an important aspect if concave contours are to be tracked.

Results on real image sequences (tracking a toy train in a laboratory scene, tracking pedestrians in an outdoor scene) show the suitability of this approach for real-time object tracking in a closed loop between image acquisition and camera movement. The contour tracking can be done within the image frame rate (25 fps) on standard Unix workstations (HP 735) without any specialized hardware.

References (0)

Cited by (46)

Regularizing deep networks with label geometry for accurate object localization on small training datasets
2022, Pattern Recognition Letters
Citation Excerpt :
It is intuitive to integrate the learned geometry into the regression branch to improve localization performance. Inspired by Active Contour Models [20] and its variant of active rays [21,22] which apply constraints of image geometry such as gradients to iterative parameter update, we couple regressed boxes with intermediate masks in a similar evolutionary manner to incorporate geometric regularization into regression. We construct an energy function on predicted distance transforms, and evolve regressed boxes via minimizing the energy function, the minimum of which defines the final sides of bounding boxes.
Localization is a critical subtask in object detection, which is closely related to spatial information of objects. Most current detectors simply rely on the fitting ability of deep neural networks to regress towards numerical targets such as coordinates of object boxes. Training deep networks for sufficient fitting ability requires a large number of annotations that are expensive to obtain. In this work, we fully exploit limited annotations by extracting label geometry to improve localization performance on small datasets. We generate distance transform of bounding box edges according to localization labels, with which we supervise intermediate outputs of networks pixel by pixel to reconstruct object geometry for localization. Distance transform is sensitive to box edges and provides geometric gradients flowing into boundaries. We learn such gradients to enhance geometric-aware features through a coupled training with regression, and use it to refine regressed boxes in an evolutionary manner in inference. Extensive experiments are implemented to demonstrate the effectiveness of our method. Our method can be applied in applications that required human-machine interaction, such as the driver-assistance system in autonomous driving, by providing accurate detections to assist humans in making better decisions.
Radial snakes: Comparison of segmentation methods in synthetic noisy images
2015, Expert Systems with Applications
Citation Excerpt :
One of the main advantages of this new approach is that it can substitute the derivative application as external energy. The derivative, used in the works of Denzler and Niemann (1996, 1999) is defined close to the edges and outside the object to be segmented, which makes it difficult to use when the initial snake is far from the edge. The use of the Hilbert transform is also gainfully related to Hunter, Soraghan, and McDonagh (1995) because the Gaussian mask is not needed to spread the influence of the derivative through the beams, also the influence of the Hilbertian energy is present at a distance greater than the proposed method in the reference mentioned above.
There has been a growing use of digital image processing since the 90’s. To process an image it must be transformed successively in order to extract information more easily. The first steps in an image analysis are the acquisition followed by the preprocessing to prepare the image for the next step. This step is called image segmentation which is the process of separating different regions of the image according to their properties. The segmentation process is fundamental for all image analyses, as the final result is essentially dependent on the quality of the segmentation. Highlighted among these techniques are the active contour systems, known as snakes. The active contour methods can be subdivided in two main groups: two-dimensional search (traditional) and one-dimensional search (radial). The radial active contours were developed in order to obtain a smaller computational cost. The aim of this work was to study, evaluate and compare algorithms of radial active contours in synthetic noisy images and thus identify the advantages and disadvantages of each method in order to point out the most appropriate method for a given application. This work makes a quantitative and qualitative comparison of three methods: Traditional Radial Snakes, Hilbert Radial Snakes and pSnakes. The results of this research are suitable for academic research as they show that the recently developed pSnakes method is effective in image segmentation with noise. This paper also considered the processing time of the different methods.
PSnakes: A new radial active contour model and its application in the segmentation of the left ventricle from echocardiographic images
2014, Computer Methods and Programs in Biomedicine
Active contours are image segmentation methods that minimize the total energy of the contour to be segmented. Among the active contour methods, the radial methods have lower computational complexity and can be applied in real time. This work aims to present a new radial active contour technique, called pSnakes, using the 1D Hilbert transform as external energy. The pSnakes method is based on the fact that the beams in ultrasound equipment diverge from a single point of the probe, thus enabling the use of polar coordinates in the segmentation. The control points or nodes of the active contour are obtained in pairs and are called twin nodes. The internal energies as well as the external one, Hilbertian energy, are redefined. The results showed that pSnakes can be used in image segmentation of short-axis echocardiogram images and that they were effective in image segmentation of the left ventricle. The echo-cardiologist's golden standard showed that the pSnakes was the best method when compared with other methods. The main contributions of this work are the use of pSnakes and Hilbertian energy, as the external energy, in image segmentation. The Hilbertian energy is calculated by the 1D Hilbert transform. Compared with traditional methods, the pSnakes method is more suitable for ultrasound images because it is not affected by variations in image contrast, such as noise. The experimental results obtained by the left ventricle segmentation of echocardiographic images demonstrated the advantages of the proposed model. The results presented in this paper are justified due to an improved performance of the Hilbert energy in the presence of speckle noise.
Size-adapted microcalcification segmentation in mammography utilizing scale-space signatures
2010, Computerized Medical Imaging and Graphics
The purpose of this study is size-adapted segmentation of individual microcalcifications in mammography, based on microcalcification scale-space signature estimation, enabling robust scale selection for initialization of multiscale active contours. Segmentation accuracy was evaluated by the area overlap measure, by comparing the proposed method and two recently proposed ones to expert manual delineations. The method achieved area overlap of 0.61 ± 0.15 outperforming statistically (p < 0.001) the other two methods (0.53 ± 0.18, 0.42 ± 0.16). Only the proposed method performed equally for both small (<460 μm) and large (≥460 μm) microcalcifications. Results indicate an accurate method, which could be utilized in computer-aided diagnosis schemes of microcalcification clusters.
HYBRID FEATURE EMBEDDING FOR AUTOMATIC BUILDING OUTLINE EXTRACTION
2023, arXiv
Hybrid Feature Embedding for Automatic Building Outline Extraction
2023, International Geoscience and Remote Sensing Symposium (IGARSS)

View all citing articles on Scopus

View full text

Regular ArticleActive Rays: Polar-transformed Active Contours for Real-Time Contour Tracking

Abstract

Regular Article
Active Rays: Polar-transformed Active Contours for Real-Time Contour Tracking