Confusion matrix is a useful and comprehensive presentation of the classifier performance. It is commonly used in the evaluation of multi-class, single-label classification models, where each data instance can belong to just one class at any given point in time. However, the real world is rarely unambiguous and hard classification of data instance to a single class, i.e. defining its properties with single distinctive feature, is not always possible. For example, an image can contain multiple objects and regions which makes multi-class classification inappropriate to describe its content. Proposed solutions to this set of problems are based on multi-label classification model where each data instance is assigned one or more labels describing its features. While most of the evaluation measures used to evaluate single-label classifier can be adapted to a multi-label classification model, presentation and evaluation of the obtained results using standard confusion matrices cannot be expanded to this case.In this paper we propose a novel method for the computation of a confusion matrix for multi-label classification. The proposed algorithm overcomes the limitations of the existing approaches in modeling relations between the classifier output and the Ground Truth (i.e. hand-labeled) classification, and due to its versatility can be used in many different research fields.

Automatic analysis of retinal fundus images is becoming increasingly present today, and diseases such as diabetic retinopathy and age-related macular degeneration are getting a higher chance of being discovered in the early stages of their development. In order to focus on discovering those diseases, researchers commonly preprocess retinal fundus images in order to detect the retinal landmarks -blood vessels, fovea and the optic disk. A large number of methods for the automatic detection of retinal blood vessels from retinal fundus images already exists, but many of them are using unnecessarily complicated approaches. In this paper we demonstrate that a reliable retinal blood vessel segmentation can be achieved with a cascade of very simple image processing methods. The proposed method puts higher emphasis on high specificity (i.e. high probability that the segmented pixels actually belong to retinal blood vessels and are not false positive detections) rather than on high sensitivity. The proposed method is based on heuristically determined parametric edge detection and shape analysis, and is evaluated on the publicly available DRIVE and STARE datasets on which it achieved the average accuracy of 96.33% and 96.10%, respectively.

In this paper, we describe a method for the prediction of concentration of chlorophyll-a (Chl-a) from satellite data in the coastal waters of Kaštela Bay and the Brač Channel (our case study areas) in the Republic of Croatia. Chl-a is one of the parameters that indicates water quality and that can be measured by in situ measurements or approximated as an optical parameter with remote sensing. Remote sensing products for monitoring Chl-a are mostly based on the ocean and open sea monitoring and are not accurate for coastal waters. In this paper, we propose a method for remote sensing monitoring that is locally tailored to suit the focused area. This method is based on a data set constructed by merging Sentinel 2 Level-2A satellite data with in situ Chl-a measurements. We augmented the data set horizontally by transforming the original feature set, and vertically by adding synthesized zero measurements for locations without Chl-a. By transforming features, we were able to achieve a sophisticated model that predicts Chl-a from combinations of features representing transformed bands. Multiple Linear Regression equation was derived to calculate Chl-a concentration and evaluated quantitatively and qualitatively. Quantitative evaluation resulted in R2 scores 0.685 and 0.659 for train and test part of data set, respectively. A map of Chl-a of the case study area was generated with our model for the dates of the known incidents of algae blooms. The results that we obtained are discussed in this paper.

No abstract

Wildfires are natural risk phenomena that cause significant economic and environmental damage. In wildfire fighting strategy it is important to detect the wildfire in its initial stage and to apply, as soon as possible, the most appropriate fire fighting action. In both cases wildfire monitoring and surveillance systems are of great importance, so in the last decade the interest for various wildfire monitoring and surveillance systems has increased, both on the research and the implementation level. This paper describes one such system named iForestFire. It is an example of advanced terrestrial vision based wildfire monitoring and surveillance system, today widely used in various Croatian National and Nature Parks and regions, but it is also a system in constant development and improvement, both on theoretical and practical level. This paper describes its last improvements in video detection part that are based on notation of observer, cogent confabulation theory and mechanism of thought. Inclusion of cogent confabulation theory allows us to expend the use of existing wildfire observers to more general natural risk observers.