In this paper some of the classic alternatives for edge detection in digital images are studied. The main idea of edge detection algorithms is to find where abrupt changes in the intensity of an image have occurred. The first family of algorithms reviewed in this work uses the first derivative to find the changes of intensity, such as Sobel, Prewitt and Roberts. In the second reviewed family, second derivatives are used, for example in algorithms like Marr-Hildreth and Haralick. The obtained results are analyzed from a qualitative point of view (perceptual) and from a quantitative point of view (number of operations, execution time), considering different ways to convolve an image with a kernel (step required in some of the algorithms). Source CodeFor all the reviewed algorithms, an open source C implementation is provided which can be downloaded from the IPOL web page of this article 1 . An online demonstration is also available, where the user can test and reproduce our results.

We present a novel segmentation algorithm based on a hierarchical representation of images. The main contribution of this work is to explore the capabilities of the a contrario reasoning when applied to the segmentation problem, and to overcome the limitations of current algorithms within that framework. This exploratory approach has three main goals.Our first goal is to extend the search space of greedy merging algorithms to the set of all partitions spanned by a certain hierarchy, and to cast the segmentation as a selection problem within this space. In this way we increase the number of tested partitions and thus we potentially improve the segmentation results. In addition, this space is considerably smaller than the space of all possible partitions, thus we still keep the complexity controlled.Our second goal aims to improve the locality of region merging algorithms, which usually merge pairs of neighboring regions. In this work, we overcome this limitation by introducing a validation procedure for complete partitions, rather than for pairs of regions.The third goal is to perform an exhaustive experimental evaluation methodology in order to provide reproducible results.Finally, we embed the selection process on a statistical a contrario framework which allows us to have only one free parameter related to the desired scale.

PurposeIn times when digitized and blended learning paradigms are getting more profuse, the COVID-19 pandemic substantially changed the dynamics of this program, forcing all the courses to migrate to virtual modality. This study highlights the biological engineering courses at the University of the Republic (Universidad de la República) in Uruguay pertaining to the adaptations to virtual learning environments during the COVID-19 pandemic and analyzing its impact through the courses taught in the virtual setting.Design/methodology/approachGlobal education has seen a significant paradigm shift over the last few years, changing from a specialized approach to a broader transdisciplinary approach. Especially in life sciences, different fields of specializations have started to share a common space in the area of applied research and development. Based on this transdisciplinary approach, the Biological Engineering program was designed at the University of the Republic (Universidad de la República), Uruguay.FindingsThe new challenges posed by the virtual modality on the pedagogical areas like course design, teaching methodologies and evaluations and logistical aspects like laboratory-setting have sparked a considerable change in different aspects of the courses. However, despite the changes to virtual modality in this year, the student-performance showed an overall improvement compared to the last year.Originality/valueWith the changing direction of pedagogy and research in biological engineering across the world, it is quintessential to adapt university courses to the same, promoting an environment where the scientific and engineering disciplines merge and the learning methodologies lead to a dynamic and adaptive ubiquitous learning environment.

Hierarchies are a powerful tool for image segmentation, they produce a multiscale representation which allows to design robust algorithms and can be stored in tree-like structures which provide an efficient implementation. These hierarchies are usually constructed explicitly or implicitly by means of region merging algorithms. These algorithms obtain the segmentation from the hierarchy by either using a greedy merging order or by cutting the hierarchy at a fixed scale.Our main contribution is to enlarge the search space of these algorithms to the set of all possible partitions spanned by a certain hierarchy, and to cast the segmentation as a selection problem within this space. The importance of this is two-fold. First, we are enlarging the search space of classic greedy algorithms and thus potentially improving the segmentation results. Second, this space is considerably smaller than the space of all possible partitions, thus we are reducing the complexity.In addition, we embed the selection process on a statistical a contrario framework which allows us to reduce the number of free parameters of our algorithm to only one.Index Terms-Image segmentation, Hierarchical systems, Statistics I. INTRODUCTIONImage segmentation is one of the oldest and most challenging problems in image processing. Given an image, even for a human observer, it is hard to determine an unique partition of the image, and it is even harder to find consensus between different observers.A usual approach taken to overcome this difficulty is to find a hierarchy of segmentations rather than an unique partition. These hierarchies are usually constructed in a bottom-up fashion, by using region merging algorithms.The first merging algorithms presented in the literature (see [1]) aimed to consecutively merge adjacent regions until a stopping criterion was met, thus yielding a single partition. All these algorithms have three basic ingredients: a region model, which tells us how to describe regions; a merging criterion, which tells us if two regions are to be merged or not; and a merging order, which tells us at each step which couple of regions should be merged first. Many of the existing approaches present similar problems: they use only region information (no boundaries are taken into account), they have a considerable number of manually tuned parameters, they use very simple region models, or they use a very simple merging order. Most of the recent work in this area has been directed to improve the merging criterion and the region model, but little effort was carried towards the merging order and the reduction of the number of parameters.