Cláudio Rosito Jung

Crowd Analysis Using Computer Vision Techniques

This article presents a survey on crowd analysis using computer vision techniques, covering different aspects such as people tracking, crowd density estimation, event detection, validation, and simulation. It also reports how related the areas of computer vision and computer graphics should be to deal with current challenges in crowd analysis.

Background Subtraction and Shadow Detection in Grayscale Video Sequences

Tracking moving objects in video sequence is an important problem in computer vision, with applications several fields, such as video surveillance and target tracking. Most techniques reported in the literature use background subtraction techniques to obtain foreground objects, and apply shadow detection algorithms exploring spectral information of the images to retrieve only valid moving objects. In this paper, we propose a small improvement to an existing background model, and incorporate a novel technique for shadow detection in grayscale video sequences. The proposed algorithm works well for both indoor and outdoor sequences, and does not require the use of color cameras.

Rectangle detection based on a windowed Hough transform

The problem of detecting rectangular structures in images arises in many applications, from building extraction in aerial images to particle detection in cryo-electron microscopy. This paper proposes a new technique for rectangle detection using a windowed Hough transform. Every pixel of the image is scanned, and a sliding window is used to compute the Hough transform of small regions of the image. Peaks of the Hough image (which correspond to line segments) are then extracted, and a rectangle is detected when four extracted peaks satisfy certain geometric conditions. Experimental results indicate that the proposed technique produced promising results for both synthetic and natural images.

Lane following and lane departure using a linear-parabolic model

This paper proposes a technique for unwanted lane departure detection. Initially, lane boundaries are detected using a combination of the edge distribution function and a modified Hough transform. In the tracking stage, a linear-parabolic lane model is used: in the near vision field, a linear model is used to obtain robust information about lane orientation; in the far field, a quadratic function is used, so that curved parts of the road can be efficiently tracked. For lane departure detection, orientations of both lane boundaries are used to compute a lane departure measure at each frame, and an alarm is triggered when such measure exceeds a threshold. Experimental results indicate that the proposed system can fit lane boundaries in the presence of several image artifacts, such as sparse shadows, lighting changes and bad conditions of road painting, being able to detect in advance involuntary lane crossings.

A lane departure warning system based on a linear-parabolic lane model

We propose a new lane departure warning system based on a linear-parabolic lane boundary model. A linear function is used to fit the near vision field, and a quadratic function fits the far field. The linear part of the model provides robust information about the orientation of the vehicle with respect to both lane boundaries, while the parabolic part is flexible enough to fit curved parts of the road. The orientation of both lane boundaries is then computed and used to anticipate lane crossings.

Efficient Background Subtraction and Shadow Removal for Monochromatic Video Sequences

This letter presents a new method for background subtraction and shadow removal for grayscale video sequences. The background image is modeled using robust statistical descriptors, and a noise estimate is obtained. Foreground pixels are extracted, and a statistical approach combined with geometrical constraints are adopted to detect and remove shadows.

A robust linear-parabolic model for lane following

In this paper we address the problem of lane detection and lane tracking. A linear model is used to approximate lane boundaries in the first frame of a video sequence, using a combination of the edge distribution function and the Hough transform. A new linear-parabolic model is used in the subsequent frames: the linear part of the model is used to fit the near vision field, while the parabolic model fits the far field. The proposed technique demands low computational power and memory requirements, and showed to be robust in the presence of noise, shadows, lack of lane painting and change of illumination conditions.

Adaptive image denoising using scale and space consistency

This paper proposes a new method for image denoising with edge preservation, based on image multiresolution decomposition by a redundant wavelet transform. In our approach, edges are implicitly located and preserved in the wavelet domain, whilst image noise is filtered out. At each resolution level, the image edges are estimated by gradient magnitudes (obtained from the wavelet coefficients), which are modeled probabilistically, and a shrinkage function is assembled based on the model obtained. Joint use of space and scale consistency is applied for better preservation of edges. The shrinkage functions are combined to preserve edges that appear simultaneously at several resolutions, and geometric constraints are applied to preserve edges that are not isolated. The proposed technique produces a filtered version of the original image, where homogeneous regions appear separated by well-defined edges. Possible applications include image presegmentation, and image denoising.

Denoising and enhancing digital mammographic images for visual screening

Dense regions in digital mammographic images are usually noisy and have low contrast, and their visual screening is difficult. This paper describes a new method for mammographic image noise suppression and enhancement, which can be effective particularly for screening image dense regions. Initially, the image is preprocessed to improve its local contrast and the discrimination of subtle details. Next, image noise suppression and edge enhancement are performed based on the wavelet transform. At each resolution, coefficients associated with noise are modelled by Gaussian random variables; coefficients associated with edges are modelled by Generalized Laplacian random variables, and a shrinkage function is assembled based on posterior probabilities. The shrinkage functions at consecutive scales are combined, and then applied to the wavelets coefficients. Given a resolution of analysis, the image denoising process is adaptive (i.e. does not require further parameter adjustments), and the selection of a gain factor provides the desired detail enhancement. The enhancement function was designed to avoid introducing artifacts in the enhancement process, which is essential in mammographic image analysis. Our preliminary results indicate that our method allows to enhance local contrast, and detect microcalcifications and other suspicious structures in situations where their detection would be difficult otherwise. Compared to other approaches, our method requires less parameter adjustments by the user.

Unsupervised multiscale segmentation of color images

This paper proposes a new multiresolution technique for color image representation and segmentation, particularly suited for noisy images. A decimated wavelet transform is initially applied to each color channel of the image, and a multiresolution representation is built up to a selected scale 2^J. Color gradient magnitudes are computed at the coarsest scale 2^J, and an adaptive threshold is used to remove spurious responses. An initial segmentation is then computed by applying the watershed transform to thresholded magnitudes, and this initial segmentation is projected to finer resolutions using inverse wavelet transforms and contour refinements, until the full resolution 2^0 is achieved. Finally, a region merging technique is applied to combine adjacent regions with similar colors. Experimental results show that the proposed technique produces results comparable to other state-of-the-art algorithms for natural images, and performs better for noisy images.