Automatic segmentation of organs and regions in medical imaging is a valuable tool for specialists. This study explores various automatic methods for lung segmentation in X-ray images. First, various neural network architectures are applied for this segmentation task, and subsequently, they are ranked based on their performance through statistical analysis. Then, as some architectures are more suitable for segmenting certain structures or regions of the X-ray, aggregation and consensus methods are studied to fuse the various neural network segmentations, with the aim of obtaining a more complete segmentation. The study reveals that the method based on the WOWA aggregation function, coupled with a maximum-based consensus method, statistically outperforms the individual segmentation provided by the best-performing neural network.

Abstract This study presents a method to improve state-of-the-art concave point detection methods as the first step towards effectively segmenting overlapping objects in images. The approach relies on analysing the curvature of the object contour. This method comprises three main steps. First, the original image is preprocessed to obtain the curvature value at each contour point. Second, the regions with higher curvatures are selected and a recursive algorithm is applied to refine previously selected regions. Finally, a concave point is obtained for each region by analysing the relative position of their neighbourhood. Furthermore, the experimental results indicate that improving the detection of concave points leads to better division of clusters. To evaluate the quality of the concave point detection algorithm, a synthetic dataset was constructed to simulate the presence of overlapping objects. This dataset includes the precise location of concave points, which serve as the ground truth for evaluation. As a case study, the performance of a well-known application, such as the splitting of overlapping cells in images of peripheral blood smears samples from patients with sickle cell anaemia, was evaluated. We used the proposed method to detect concave points in cell clusters and then separated these clusters by ellipse fitting.

Patients undergoing abdominal operations should perform a postoperative follow-up. The subsequent face-to-face consultations have an economic cost to the health system and can affect the quality of life of patients. The follow-up of the patients could be simplified using a smartphone application that, based on an image of the surgical wound and a simple questionnaire, could detect postoperative complications. One of the main complications is the appearance of infection in the wound, signaled by the redness that appears around the wound and the staples. In this paper we present a method to automatically detect infection in surgical wounds. The method has been divided into three steps. First step in this process is to locate and remove the staples that keep the wound closed because they distort the overall color of the image. To do this, after transforming the image to the HSV color space, we apply the Sobel gradient in the vertical direction of the V channel. The discrete fuzzy mathematical morphology process it to yield a binary mask with areas indicating the position of the staples. Second, using a fuzzy mathematical morphology for images in the CIELab color space, we apply an inpainting algorithm to restore the regions detected as staples. Third, using triangular and trapezoidal fuzzy sets, we determine if there are areas of the wound that present a deviation to the red color. Our experiments show that the proposed method can be successfully used to detect the possibility of infection in the wound. In addition, the algorithms have been designed using simple operations so that the method could be introduced in a smartphone application.

Morphological operators have been used extensively when dealing with binary or grayscale images, but there is no general-purpose approach for multivariate images that meets the expectactions of practitioners under diferent circumstances. Although several approaches have been proposed, state-of-the-art applications tend to process channels independently, obviating interchannel correlation. In this work, we introduce a new definition of erosion and dilation that can handle images with any number of channels, study their theoretical properties and analyse its behaviour. It is based on the Fuzzy Mathematical Morphology, from which it inherits essential theoretical properties. Our operators consider the first channel to evaluate a pixel's importance, but handles all channels to generate coherent outputs. It successfully processes natural images in the L*a*b* space and can also avoid the creation of new chromatic values, specially important for hyperspectral imagery. We provide, thus, a general and well-founded framework to process color images with morphological operators.

In many computer vision applications, vignetting is an undesirable effect which must be removed in a pre-processing step. Recently, an algorithm for image vignetting correction has been presented by means of a minimization of log-intensity entropy. This method relies on an increase of the entropy of the image when it is affected with vignetting. In this paper, we propose a novel algorithm to reduce image vignetting via a maximization of the fuzzy entropy of the image. Fuzzy entropy quantifies the fuzziness degree of a fuzzy set and its value is also modified by the presence of vignetting. The experimental results show that this novel algorithm outperforms in most cases the algorithm based on the minimization of log-intensity entropy both from the qualitative and the quantitative point of view.

The proposal of new edge detectors is a constant in last years due to the importance of the edge detection result for high-level image processing tasks. It is known that there is no optimal edge detector for all kinds of images and in addition, it is not a straightforward mission to set the optimal set of parameters of the edge detector for each image. In this paper, an algorithm to deal with these drawbacks is presented based on obtaining a consensus edge image from the edge images obtained by several edge detectors or configurations of the same edge detector. These input edge images are combined using multifuzzy sets and averaging aggregation functions and after applying a penalty function, the consensus edge image is obtained. Experimental results show that this approach leads to a robust edge detector which is able to obtain notable results for different kinds of images.

Noise reduction is a fundamental step in many image processing and computer vision applications. In recent years, several noise filters especially devoted to the removal of high density salt and pepper noise, as a particular case of impulse noise, have been proposed. In this paper, a novel two-stage filter based on the fuzzy mathematical morphology using t-norms and aggregation functions is proposed. This filter involves a detection step of the noisy pixels and the restoration of the image by means of the aggregation of the non-corrupted pixels through an adequate weighted arithmetic mean. The experimental results show that the proposed algorithm outperforms other filtering methods both from the visual point of view and the values of some objective performance measures for images corrupted from a 25% up to 98% of noise.

In this paper, the fuzzy morphological gradients from the fuzzy mathematical morphologies based on t-norms and conjunctive uninorms are deeply analyzed in order to establish which pair of conjunction and fuzzy implications are optimal, in accordance with their performance in edge detection applications. A novel three-step algorithm based on the fuzzy morphology is proposed. The comparison is performed by means of the so-called Pratt's figure of merit. In addition, a statistical analysis is carried out to study the relationship between the different configurations and to establish a classification of the conjunctions and implications considered. Both the objective measure and the statistical analysis conclude that the pairs nilpotent minimum t-norm and the Kleene-Dienes implication, and the idempotent uninorm obtained with the classical negation as a generator and its residual implication, are the best configurations in this approach, because they also obtain competitive results with respect to other approaches.

The study of cell morphology is an important aspect of the diagnosis of some diseases, such as sickle cell disease, because red blood cell deformation is caused by these diseases. Due to the elongated shape of the erythrocyte, ellipse adjustment and concave point detection are applied widely to images of peripheral blood samples, including during the detection of cells that are partially occluded in the clusters generated by the sample preparation process. In the present study, we propose a method for the analysis of the shape of erythrocytes in peripheral blood smear samples of sickle cell disease, which uses ellipse adjustments and a new algorithm for detecting notable points. Furthermore, we apply a set of constraints that allow the elimination of significant image preprocessing steps proposed in previous studies. We used three types of images to validate our method: artificial images, which were automatically generated in a random manner using a computer code; real images from peripheral blood smear sample images that contained normal and elongated erythrocytes; and synthetic images generated from real isolated cells. Using the proposed method, the efficiency of detecting the two types of objects in the three image types exceeded 99.00%, 98.00%, and 99.35%, respectively. These efficiency levels were superior to the results obtained with previously proposed methods using the same database, which is available at http://erythrocytesidb.uib.es/. This method can be extended to clusters of several cells and it requires no user inputs.

In this paper, a new approach to fuzzy mathematical morphology based on discrete t-norms is studied. It is proved that the most usual algebraic and morphological properties are preserved, such as, duality, monotonicity, interaction with union and intersection, invariance under translating and scaling, local knowledge property, extensitivity, idempotence, and many others. In fact, all properties satisfied by the approach based on nilpotent t-norms hold in the discrete case. This is quite important since in practice we only work with discrete objects. Experimental results for some discrete t-norms are included. They are compared with classical morphological algorithms based on the Łukasiewicz t-norm and the umbra approach, and with the fuzzy approach based on idempotent uninorms, proving that they are suitable to be used in edge detection.

Medical images edge detection is one of the most important pre-processing steps in medical image segmentation and 3D reconstruction. In this paper, an edge detection algorithm using an uninorm-based fuzzy morphology is proposed. It is shown that this algorithm is robust when it is applied to different types of noisy images. It improves the results of other well-known algorithms including classical algorithms of edge detection, as well as fuzzy-morphology based ones using the {\L}ukasiewicz t-norm and umbra approach. It detects detailed edge features and thin edges of medical images corrupted by impulse or gaussian noise. Moreover, some different objective measures have been used to evaluate the filtered results obtaining for our approach better values than for other approaches.

Due to the fact that the aging human face encompasses skull bones, facial muscles, and tissues, we render it using the effects of flaccidity through the observation of family groups categorized by sex, race and age. Considering that patterns of aging are consistent, facial ptosis becomes manifest toward the end of the fourth decade. In order to simulate facial aging according to these patterns, we used surfaces with control points so that it was possible to represent the effect of aging through flaccidity. The main use of these surfaces is to simulate flaccidity and aging consequently.

In many applications the study of human movement using a computer vision and graphics techniques is very useful. One of these applications is the three-dimensional reconstruction of the structure of the human body and its movement using sequences of images and biomechanical graphics models. We present a whole and general system to study the human motion without markers but, in this case, we apply it to high-level competition disabled swimmers. This kind of study needs accuracy on the analysis and reconstruction of the person's body, therefore the virtual human (avatar) must have similar anthropometrical characteristics than the person who is doing the movement. We define a process to adjust the humanoid to the morphology of the person. It could be very laborious and subjective if done manually or by selection of points, but in this article we present a global human motion system capturing, modeling and matching a semiautomatic process between the real person and the modeled humanoid or synthetic avatar. Semiautomatic means that the computers propose the best matching from previous frames and the user can accept it or not. Once this process is carried out we are ready to analyze and represent the movements under study. It must be defined a specific set of parameter for every kind of sport. The study is adapted to specific sport activities, in our case swimming activities with an additional complexity: water occlusion and distortion. In these cases the adjustment process must be assisted by the computer using rules and models to help the expert user in correspondence tasks.