Eduard Sojka

A new algorithm for detecting corners in digital images

Corners and vertices are important features in images, which are frequently used for scene analysis, stereo matching and object tracking. Many algorithms for detecting the corners have been reported up to now. In this paper, a new algorithm is presented. The algorithm is based on measuring the variance of the directions of the gradient of brightness. The probability of the event that a point belongs to the approximation of a straight segment of the isoline of brightness passing through the point being tested is computed using the technique of Bayesian estimations and used as a weight. The algorithm is presented both from the theoretical and from the practical point of view. The results of tests show that the algorithm is better than other known algorithms that are usually used for solving the problem.

A New and Efficient Algorithm for Detecting the Corners in Digital Images

The corners and vertices are important features in images, which are frequently used for scene analysis, stereo matching and object tracking. Many algorithms for detecting the corners have been developed up to now. In this paper, a new and efficient algorithm is presented. The probability of the event that a point belongs to the approximation of the straight segment of the isoline of brightness containing the corner candidate being tested is determined using the technique of Bayesian estimations, and then used for computing the angle between the segments. The results of the tests show that, in the sense of the successfulness of detection, the new algorithm is better than the other known algorithms that are usually used for solving the problem.

A new approach to detecting the corners in digital images

The corners are important features in images that are frequently used for scene analysis, image matching and object tracking. In this paper, a new corner detector is presented that is based on a theory that takes into account relatively complicated situations that may occur in the neighbourhoods of the corner candidates. The new detector was implemented, tested, and compared with several other existing algorithms that are often used to solve the problem. In the sense of successfulness and stability of detection, the new algorithm was better in the tests than the other algorithms that were used for comparison. The ANSI C code of the detector is available to the public.

A Motion Estimation Method Based on Possibility Theory

In this paper, we present a new technique for determining the motion field in the image sequences that is based on possibility theory and that provides a good trade-off between the accuracy and speed of computation. We present the theoretical background of the method, the description of the corresponding algorithm, and experimental results. The method may be regarded as accurate and fast at the same time.

Hierarchical blurring mean-shift

In recent years, various Mean-Shift methods were used for filtration and segmentation of images and other datasets. These methods achieve good segmentation results, but the computational speed is sometimes very low, especially for big images and some specific settings. In this paper, we propose an improved segmentation method that we call Hierarchical Blurring Mean-Shift. The method achieve significant reduction of computation time and minimal influence on segmentation quality. A comparison of our method with traditional Blurring Mean-Shift and Hierarchical Mean-Shift with respect to the quality of segmentation and computational time is demonstrated. Furthermore, we study the influence of parameter settings in various hierarchy depths on computational time and number of segments. Finally, the results promising reliable and fast image segmentation are presented.

AdaBoost for Parking Lot Occupation Detection

Parking lot occupation detection using vision systems is a very important task. Many systems use different sensors and their combinations to find out whether the parking lot or space is occupied or not. Using CCTV systems makes it possible to monitor great areas without a need of many sensors. In this paper, we present a method that uses the boosting algorithm for car detection on particular parking spaces and shifting the image to obtain a probability function of car appearance. Using the model of parking lot, we achieve occupancy of each parking space. We also experimented with the detector that is based on the histogram of oriented gradients (HOG) with a support vector machine (SVM). Nevertheless, we found some drawbacks of this detector that we describe in experiments. On the grounds of these drawbacks, we decided to use the AdaBoost based detector.

Energy-transfer features and their application in the task of face detection

In this paper, we describe a novel and interesting approach for extracting the image features. The features we propose are efficient and robust; the feature vectors of relatively small dimensions are sufficient for successful recognition. We call them the energy-transfer features. In contrast, the classical features (e.g. HOG, Haar features) that are combined with the trainable classifiers (e.g. a support vector machine, neural network) require large training sets due to their high dimensionality. The large training sets are difficult to acquire in many cases. In addition to that, the large training sets slow down the training phase. Moreover, the high dimension of feature vector also slows down the detection phase and the methods for the reduction of feature vector must be used. These shortcomings became the motivation for creating the features that are able to describe the object of interest with a relatively small number of numerical values without the use of methods for the reduction of feature vector. In this paper, we demonstrate the properties of our features in the task of face detection.

Active Contours without Edges and with Simple Shape Priors

In this paper, we introduce two simple shape priors into the Chan and Vese level-set method, namely, a prescribed area and a prescribed area-to-perimeter ratio of particular objects. It is remarkable that these priors may be easily incorporated into the Euler-Lagrange equation of the original method. As a side effect of our experimenting with the method, we also introduce a new probability-based level-set function, which removes the need for reinitialisation and usually, according to our experience, speeds up the computation. Finally, we also propose a method how to treat, in a simple way, the situation in which the particular objects differ in brightness. Although the mentioned changes make the segmentation more reliable, they almost do not complicate the computation. The results of experiments are also presented.

Energy transfer features combined with DCT for object detection

The basic idea behind the energy transfer features is that the appearance of objects can be described using a function of energy distribution in images. Inside the image, the energy sources are placed and the energy is transferred from the sources during a certain chosen time. The values of energy distribution function have to be reduced into a reasonable number of values. The process of reducing can be simply solved by sampling. The input image is divided into regular cells. The mean value is calculated inside each cell. The values of samples are then considered as a vector that is used as an input for the SVM classifier. We propose an improvement to this process. The discrete cosine transform coefficients are calculated inside the cells (instead of the mean values) to construct the feature vector for the face and pedestrian detectors. To reduce the number of coefficients, we use the patterns in which the coefficients are grouped into regions. In the face detector, the principal component analysis is also used to create the feature vector with a relatively small dimension. The results show that, using this approach, the objects can be efficiently encoded with a relatively short vector with the results that outperform the results of the state-of-the-art detectors.

Resistance-Geodesic Distance and Its Use in Image Segmentation

Measuring the distance is an important task in many clustering and image-segmentation algorithms. The value of the distance decides whether two image points belong to a single or, respectively, to two different image segments. The Euclidean distance is used quite often. In more complicated cases, measuring the distances along the surface that is defined by the image function may be more appropriate. The geodesic distance, i.e. the shortest path in the corresponding graph, has become popular in this context. The problem is that it is determined on the basis of only one path that can be viewed as infinitely thin and that can arise accidentally as a result of imperfections in the image. Considering the k shortest paths can be regarded as an effort towards the measurement of the distance that is more reliable. The drawback remains that measuring the distance along several paths is burdened with the same problems as the original geodesic distance. Therefore, it does not guarantee significantly better results. In addition to this, the approach is computationally demanding. This paper introduces the resistance-geodesic distance with the goal to reduce the possibility of using a false accidental path for measurement. The approach can be briefly characterised in such a way that the path of a certain chosen width is sought for, which is in contrast to the geodesic distance. Firstly, the effective conductance is computed for each pair of the neighbouring nodes to determine the local width of the path that could possibly run through the arc connecting them. The width computed in this way is then used for determining the costs of arcs; the arcs whose use would lead to a small width of the final path are penalised. The usual methods for computing the shortest path in a graph are then used to compute the final distances. The corresponding theory and the experimental results are presented in this paper.