The whole world facing a huge crisis because of Corona virus also known as COVID-2019, identified first in December 2019 in the city of Wuhan located in China. The detection of persons infected with the virus is most important as it can be spread easily from him to others and also the person infected with the virus may not know that he is infected until a number of symptoms fallout from him. In this paper the virus detection is done using deep learning and machine learning algorithms using the X-ray images. A dataset is created with three classes consisting of normal, corona virus, and pneumonia images. The proposed method uses ResNet50 and SVM, deep learning features are extracted using ResNet50 and classification is done using SVM classifier. The classification accuracy obtained from the model is 100% when testing on the Corona virus and normal images, whereas the results obtained from the model is 94% when it is tested on the dataset consisting of normal, Corona virus and pneumonia images and performed well compared to VGG16.
Face detection has become an essential biometric method used in unlocking the smart mobiles now-a-days. Face detection identifies the face of the user by calculating the facial features. In some cases, people can use the photos and face masks to hack mobile security systems. In order to avoid this eye blinking detection, which finds eyes through the proportion of human face was proposed in this paper. The Proposed method detects the movements of eyeball and the number of eye blinking to improve face recognition for screen unlock on the mobile devices. To achieve this halfway face acknowledgment approach by adjusting the test fractional face to exhibition faces utilizing the geometrical and textural data of the extricated nearby highlights was proposed in this model to obtain better results. To extend further a hearty strategy called stepwise direct discriminant examination centers around choosing the limited highlights from the face outlines and separating their class dependent on relapse esteems was done. The motivation behind an element extraction system is to remove the confined highlights from appearances that the past component extraction procedures were constrained in examining. Our essential instinct is that if the test halfway face fix and the exhibition face picture are from a similar individual, the cost capacity of our arrangement strategy ought to be limited. Experimental results proved that our method is efficient and robust for the screen unlock on the mobile devices.
25 December 2018 in Indonesian Journal of Electrical Engineering and Informatics (IJEEI)
Nowadays, huge data bases are required to store the Digital medical images so that they can be accessed easily on requirement. To retrieve the diagnostic images, radiologist and physicians are using Content based image retrieval (CBIR). Algorithms extract features like texture, edge, color and shape from an image in CBIR systems and these extracted features from the input and are compared for similarity with the features of images in database. In this paper, Lossless compression is used for storage and effective transmission in inadequate bandwidth. Visually lossless image compression is obtained using the Daubechies wavelet with Huffman coding. Gabor transforms are utilized to extract the shape and texture features from the images. Features are selected with Mutual Information (MI) and the proposed wrapper based Cuckoo Search (CS) technique. Extracted features are fed as input to the proposed partial Recurrent Neural Networks (RNN) for the classification. The network is optimized hybrid Particle Swarm Optimization and Cuckoo Search. It was observed that the classification accuracy acquired is satisfactory
This paper discusses an application of multilayer Perceptron in wireless sensor network security. The wireless sensor network (WSN) is a network of distributed autonomous devices that supervise physical or environmental conditions corporately. The multilayer perceptron (MLP) based media access control protocol (MAC) to secure CSMA-based wireless sensor networks against denial-of-service attacks launched by adversaries. The MLP boost security of a WSN by consistently monitoring the parameter that reveals unusual variation in case of an attack the MLP alerts the MAC layer and the physical layer of the sensor mode when the suspicion factor, the output of the MLP exceeds a preset threshold level. Back propagation and Radial Basis Function Network (RBFN) algorithms are used for training the MLP. Simulation results show that the MLP helps in extending the life time of the WSN.