Narrow Results

COVID-19 is a challenging worldwide pandemic disease nowadays that spreads from person to person in a very fast manner. It is necessary to develop an automated technique for COVID-19 identification. This work investigates a new framework that predicts COVID-19 based on X-ray images. The suggested methodology contains core phases as preprocessing, feature extraction, selection and categorization. The Guided and 2D Gaussian filters are utilized for image improvement as a preprocessing phase. The outcome is then passed to 2D-superpixel method for region of interest (ROI). The pre-trained models such as Darknet-53 and Densenet-201 are then applied for features extraction from the segmented images. The entropy coded GLEO features selection is based on the extracted and selected features, and ensemble serially to produce a single feature vector. The single vector is finally supplied as an input to the variations of the SVM classifier for the categorization of the normal/abnormal (COVID-19) X-rays images. The presented approach is evaluated with different measures known as accuracy, recall, F1 Score, and precision. The integrated framework for the proposed system achieves the acceptable accuracies on the SVM Classifiers, which authenticate the proposed approach’s effectiveness.

Edge detection is a major step in several computer vision applications. Edges define the shape of objects to be used in a recognition system, for example. In this work, we introduce an approach to edge detection inspired by a challenge for artists: the Speed Drawing Challenge. In this challenge, a person is asked to draw the same figure in different times (as 10min, 1min and 10s); at each time, different levels of details are drawn by the artist. In a short time stamp, just the major elements remain. This work proposes a new approach for producing images with different amounts of edges representing different levels of relevance. Our method uses superpixel to suppress image details, followed by Globalized Probability of Boundary (gPb) and Canny edge detection algorithms to create an image containing different number of edges. After that, an edge analysis step detects whose edges are the most relevant for the scene. The results are presented for the BSDS500 dataset and they are compared to other edge and contour detection algorithms by quantitative and qualitative means with very satisfactory results.

In modern society, digital images can be far-reaching, and the images are manipulated by various software and hardware technologies. The image forgery activities are undertaken by the attackers mainly for damaging the reputation of people or receiving fiscal gain, etc. Taking this into consideration, many techniques are developed to detect the forged images. In this paper, a new deep learning-based approach is introduced for copy-move forgery detection. The input images are segmented into non-overlapping patches using superpixel-based modified dense peak clustering and the features are extracted from the segmented patches by applying deep learning structure of attention-based DenseNet 121 model. Besides, to compare every block, the depth of each pixel is reconstructed, and eventually matching process is carried out using the adaptive chimp patch matching approach, which detects the suspicious forged regions in an image. Finally, the matched keypoints are merged with the segmented patches using the merged keypoint matching algorithm. As a result, the new deep learning approach has detected the forged regions efficiently from the tampered image with less time even the image is compressed, rotated, or scaled. The performance is evaluated in terms of recall, precision, accuracy, F1-score, computational time, and False Positive Rate (FPR). Moreover, the performance is compared with the other existing approaches, and the outcomes showed that the proposed method has achieved higher accuracy of 97%, recall of 99%, precision of 97.84%, F1-score of 98.81%, FPR of 0.022 and less computational time of 2.5 s.

As a powerful classifier, sparse representation-based classification (SRC) has successfully been applied in various visual recognition problems. However, due to the highly correlated bands and insufficient training samples of hyperspectral image (HSI) data, it still remains a challenging problem to effectively apply SRC in HSI. Considering the rich information of spatial structure of materials in HSI, that means the adjacent pixels belong to the same class with a high probability, in this paper, we propose an efficient superpixel-based sparse representation framework for HSI classification. Each superpixel can be regarded as a small region consisting of a number of pixels with similar spectral characteristics. The proposed framework utilizes superpixel to exploit spatial information which can greatly improve classification accuracy. Specifically, SRC is firstly used to classify the HSI data. Meanwhile, an efficient segmentation algorithm is applied to divide the HSI into many disjoint superpixels. Then, each superpixel is used to fuse the SRC classification results in superpixel level. Experimental results on two real-world HSI data sets have shown that the proposed superpixel-based SRC (SP-SRC) framework has a significant improvement over the pixel-based SRC method.