Pavel Vlašánek

University of Ostrava · Institute for Research and Applications of Fuzzy Modeling

Data shift is a gap between data distribution used for training and data distribution encountered in the real-world. Data augmentations help narrow the gap by generating new data samples, increasing data variability, and data space coverage. We present a new data augmentation: Unproportional mosaicing (Unprop). Our augmentation randomly splits an i...

We present a new version of YOLO with better performance and extended with instance segmentation called Poly-YOLO. Poly-YOLO builds on the original ideas of YOLOv3 and removes two of its weaknesses: a large amount of rewritten labels and an inefficient distribution of anchors. Poly-YOLO reduces the issues by aggregating features from a light SE-Dar...

We present a new version of YOLO with better performance and extended with instance segmentation called Poly-YOLO. Poly-YOLO builds on the original ideas of YOLOv3 and removes two of its weaknesses: a large amount of rewritten labels and inefficient distribution of anchors. Poly-YOLO reduces the issues by aggregating features from a light SE-Darkne...

We consider the problem of inpainting for relatively large damaged areas where the best-known results are achieved using patches. We stem from the ROF-type model. We propose a new ROF model with nonlocal operators and modify it with the F-transform-based operators. As a result, the minimization is considered over a searching space restricted to a f...

The paper is focused on demonstration of image inpainting technique using the F-transform theory. Side by side with many algorithms for the image reconstruction we developed a new method of patch-based filling of an unknown (damaged) image area. The unknown area is proposed to be recursively filled by those known patches that have non-empty overlap...

We propose to solve the problem of image in-painting using the F-transform technique particularly of thezero and first degrees. We focus on the so-called patchinpainting. The proposed technique creates a feature vectorof an image that characterizes the weighted average levels ofthe corresponding intensity function and its partial derivativesover ce...

In the proposed contribution, we have discussed the usefulness of the higher degree $F^p$-transforms, $p=0,1$, for various tasks of image processing. We show that an image can be efficiently represented as a matrix of its F-transform components. We analyze the details and discuss advantages of this type of representation. We show that in a particul...

We propose the technique of the semi-automatic image creation. By this we mean an automatic completion of an image that is partially defined on the given domain. The essential feature of this technique is that the complementary area is much larger than that where the image is defined. Moreover, the proposed technique can be used in image upsampling...

Discrete convolution is commonly used operation in the image processing field. The technique modifies an input image in various ways by scalable and easy computation. We are able to achieve many applications such as edge detection, noise reduction or artistic filtering, by selection of the proper kernel. The technique of the F-transform investigate...

Many interpolation techniques are available for image reconstruction, with differences in time complexity, memory complexity and quality. In this article, we compare the application of bilinear interpolation, nearest neighbor interpolation and the F-transform approximation technique to the problem of image reconstruction. Based on our results, F-tr...

We propose a new image reconstruction technique that uses approximation properties of the fuzzy (F-)transform and its ability to remove noise. The proposed technique is based on the inverse F-transform and combines it with an original image on a non-damaged area. We present two algorithms of the F-transform based reconstruction: one-step and multi-...

Technique of image reconstruction using Ftransform uses basic functions for computation. Radius and partition of these functions affects quality of the reconstruction. In this article we are focusing on influence of the basic function shape on the quality of reconstruction and providing results of nondecreasing, nonincreasing and oscillating exampl...

Image reconstruction technique based on F-transform uses clearly defined basic functions. These functions have strong impact on the quality of reconstruction. We can use some predefined shape and radius, but also we can create a new one from the scratch. The aim of this paper is to analyze the creating process and based on that find best basic func...

The article is focused on image reconstruction with usage of the F-transform. The results will be demonstrated on the picture of the Lena damaged by three kinds of the masks. We will compare reconstructed images based on the 2D F-transform with reconstructed images based on Gaussian radial basis function.

We propose a new image reconstruction technique which uses approximation properties of the F-transform and its ability to remove noise. The proposed technique is based on the inverse F-transform and on a fusion which combines the inverse F-transform with the original image on a non-damaged area. We claim the the new technique has less computation c...

The paper is focused on using the webcam as a 3D input device. Generally it means the user may use any object and shadow of this object to obtain the coordinates x, y and z (position in the space). This idea allows to use a navigation in three dimensions without an expensive or atypical hardware. The system needs one ordinary webcam, an ordinary co...