Papers by Tomasz Marciniak
This paper presents a calculation of the motion measures for intelligent video analysis. Four mea... more This paper presents a calculation of the motion measures for intelligent video analysis. Four measures were calculated: Bidirectional Length, Forward Direction, Backward Direction and Motion Edges. All measures are based on flow map, calculated with the use of the optical flow algorithm. Motion Edges utilizes Finite-Time Lyapunov Exponent (FTLE) fields. Experimental tests were performed for video recordings with the people movement. The computation times of particular steps of the proposed solution were also checked.
The paper presents an analysis of visualization methods of crowd density visualization. Generated... more The paper presents an analysis of visualization methods of crowd density visualization. Generated density maps take into account changes in time. Three methods have been implemented and tested. The first one uses motion detection based on the background subtraction. The second one is based on BLOBs (binary large objects) analysis. The third method uses interest points ie. points on the image that can be used by the object track the movement. The tests were performed using the PETS2009 video sequence database. The obtained maps were evaluated and the time consumptions were estimated.
Sensors
Optical coherence tomography (OCT) is one of the newest and most important optical non-invasive m... more Optical coherence tomography (OCT) is one of the newest and most important optical non-invasive methods for the investigation and testing of various materials (e [...]
Annals of Computer Science and Information Systems
Sensors
The use of neural networks for retinal vessel segmentation has gained significant attention in re... more The use of neural networks for retinal vessel segmentation has gained significant attention in recent years. Most of the research related to the segmentation of retinal blood vessels is based on fundus images. In this study, we examine five neural network architectures to accurately segment vessels in fundus images reconstructed from 3D OCT scan data. OCT-based fundus reconstructions are of much lower quality compared to color fundus photographs due to noise and lower and disproportionate resolutions. The fundus image reconstruction process was performed based on the segmentation of the retinal layers in B-scans. Three reconstruction variants were proposed, which were then used in the process of detecting blood vessels using neural networks. We evaluated performance using a custom dataset of 24 3D OCT scans (with manual annotations performed by an ophthalmologist) using 6-fold cross-validation and demonstrated segmentation accuracy up to 98%. Our results indicate that the use of neu...
Metrology and Measurement Systems, 2019
In this paper methods and their examination results for automatic segmentation and parameterizati... more In this paper methods and their examination results for automatic segmentation and parameterization of vessels based on spectral domain optical coherence tomography (SD-OCT) of the retina are presented. We present three strategies for morphologic image processing of a fundus image reconstructed from OCT scans. A specificity of initial image processing for fundus reconstruction is analysed. Then, the parameterization step is performed based on the vessels segmented with the proposed algorithm. The influence of various methods on the vessel segmentation and fully automatic vessel measurement is analysed. Experiments were carried out with a set of 3D OCT scans obtained from 24 eyes (12 healthy volunteers) with the use of an Avanti RTvue OCT device. The results of automatic vessel segmentation were numerically compared with those prepared manually by the medical doctor experts.
Sensors, 2021
This paper proposes an efficient segmentation of the preretinal area between the inner limiting m... more This paper proposes an efficient segmentation of the preretinal area between the inner limiting membrane (ILM) and posterior cortical vitreous (PCV) of the human eye in an image obtained with the use of optical coherence tomography (OCT). The research was carried out using a database of three-dimensional OCT imaging scans obtained with the Optovue RTVue XR Avanti device. Various types of neural networks (UNet, Attention UNet, ReLayNet, LFUNet) were tested for semantic segmentation, their effectiveness was assessed using the Dice coefficient and compared to the graph theory techniques. Improvement in segmentation efficiency was achieved through the use of relative distance maps. We also show that selecting a larger kernel size for convolutional layers can improve segmentation quality depending on the neural network model. In the case of PVC, we obtain the effectiveness reaching up to 96.35%. The proposed solution can be widely used to diagnose vitreomacular traction changes, which is...
Electronics, 2021
The aim of the paper was to analyze effective solutions for accurate lane detection on the roads.... more The aim of the paper was to analyze effective solutions for accurate lane detection on the roads. We focused on effective detection of airport runways and taxiways in order to drive a light-measurement trailer correctly. Three techniques for video-based line extracting were used for specific detection of environment conditions: (i) line detection using edge detection, Scharr mask and Hough transform, (ii) finding the optimal path using the hyperbola fitting line detection algorithm based on edge detection and (iii) detection of horizontal markings using image segmentation in the HSV color space. The developed solutions were tuned and tested with the use of embedded devices such as Raspberry Pi 4B or NVIDIA Jetson Nano.
2017 25th European Signal Processing Conference (EUSIPCO), Aug 1, 2017
In this paper we present a method for volumetric segmentation of retinal vessels based on 3D OCT ... more In this paper we present a method for volumetric segmentation of retinal vessels based on 3D OCT images of human macula. The proposed hybrid method is comprised of two steps: detailed extraction of superficial blood vessels indicators visible in 2D projection of retina layers followed by an axial inspection of inner retina to determine exact depth position of each vessel. The segmentation procedure is improved by application of block-matching and 4D filtering (BM4D) algorithm for noise reduction. The 3D reconstruction of vascular structure was performed for 10 normal subjects examined with Avanti AngioVue OCT device. The automated segmentation results were validated against the manual segmentation performed by an expert giving the accuracy of 95.2%.
Pomiary Automatyka Robotyka, 2019
Bulletin of the Polish Academy of Sciences Technical Sciences, 2017
This paper presents analysis of selected noise reduction methods used in optical coherence tomogr... more This paper presents analysis of selected noise reduction methods used in optical coherence tomography (OCT) retina images (the socalled B-scans). The tested algorithms include median and averaging filtering, anisotropic diffusion, soft wavelet thresholding, and multiframe wavelet thresholding. Precision of the denoising process was evaluated based on the results of automated retina layers segmentation, since this stage (vital for ophthalmic diagnosis) is strongly dependent on the image quality. Experiments were conducted with a set of 3D low quality scans obtained from 10 healthy patients and 10 patients with vitreoretinal pathologies. Influence of each method on the automatic image segmentation for both groups of patients is thoroughly described. Manual annotations of investigated retina layers provided by ophthalmology experts served as reference data for evaluation of the segmentation algorithm.
2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA), 2016
Retinal pathologies that are detected too late and/or left untreated can seriously damage eyesigh... more Retinal pathologies that are detected too late and/or left untreated can seriously damage eyesight. It is important to monitor the retina and react to any pathological changes. A fast, accurate, noninvasive, and even three-dimensional retina examination is the optical coherence tomography (OCT). In this paper we propose a new automated classification method for evaluation of vitreomacular interface (VRI) in human eyes. We present an approach for modelling changes in retina structure during the progression of vitreomacular traction (VMT) pathology. Presented experiments were performed on volumetric data acquired from adult patients with the use of Avanti RTvue device. Advanced digital image processing algorithms were subsequently applied to each OCT cross-section (B-scan) for image denoising and flattening, as well as retina layers segmentation. The proposed solution has a good accuracy and almost all subjects were successfully classified into one of 4 groups corresponding to various stages of VMT. The developed models of VMT stages show a high potential of the proposed method to support ophthalmologists in making appropriate clinical decisions.
2015 Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA), 2015
The paper presents implementation aspects of real time speaker recognition from Internet radio br... more The paper presents implementation aspects of real time speaker recognition from Internet radio broadcasts. The proposed solution is based on the free and open source Python development tools. The prepared software was tested with the Windows environment and then adapted to the Unix operating system running on the Raspberry Pi platform. We show what libraries are most convenient in order to implement individual blocks of the speaker recognition algorithm. In the paper we also indicate parameters, for which the algorithm exhibits the greatest efficiency. The prepared software is available on the Github file repository.
Foundations of Computing and Decision Sciences, 2015
In this paper we present an automatic speaker recognition technique with the use of the Internet ... more In this paper we present an automatic speaker recognition technique with the use of the Internet radio lossy (encoded) speech signal streams. We show an influence of the audio encoder (e.g., bitrate) on the speaker model quality. The model of each speaker was calculated with the use of the Gaussian mixture model (GMM) approach. Both the speaker recognition and the further analysis were realized with the use of short utterances to facilitate real time processing. The neighborhoods of the speaker models were analyzed with the use of the ISOMAP algorithm. The experiments were based on four 1-hour public debates with 7–8 speakers (including the moderator), acquired from the Polish radio Internet services. The presented software was developed with the MATLAB environment.
In this paper two vision systems for urban traffic improvement are presented. In both of them inf... more In this paper two vision systems for urban traffic improvement are presented. In both of them information from a camera mounted on a vehicle is used. The first system automatically detects traffic lights changes from red to green and allows to start movement of the car without unnecessary delays. The second system automatically detects turn signal of a car, which is going to change the lane. This provides information to help the proper use of the late merge scheme in situations when two lanes in one direction are merged into one lane. Initial experiments indicate high efficiency of the proposed mechanisms.
Metrology and Measurement Systems, 2016
This paper presents signal processing aspects for automatic segmentation of retinal layers of the... more This paper presents signal processing aspects for automatic segmentation of retinal layers of the human eye. The paper draws attention to the problems that occur during the computer image processing of images obtained with the use of the Spectral Domain Optical Coherence Tomography (SD OCT). Accuracy of the retinal layer segmentation for a set of typical 3D scans with a rather low quality was shown. Some possible ways to improve quality of the final results are pointed out. The experimental studies were performed using the so-called B-scans obtained with the OCT Copernicus HR device.
Uploads
Papers by Tomasz Marciniak