ملخص البحث :

This paper introduces an approach for automatic underwater object classification which consists of six main steps: noise removal and improve the appearances of the input image, then automatically detect and segment the objects based on salient object detection techniques from the image, extract the object features by combining the results of two feature descriptors, learning, classify the object and finally, measure the approach performance with five factors. This paper provides two major contributions: first, combine SIFT after normalization with PCA and GLCM; second, the classification step is based on salient object detection. There are multi models included within the proposed approach and this paper introduces comparisons between the obtained results.

ملخص البحث :

Medical image enhancement is a topic of great interest to researchers due to the rapid evolution of technology and advancements in communication. There are many types of medical images such as X-ray images, computed tomography(CT) scans, magnetic resonance imaging (MRI) scans, ultrasound images, positron emission tomography(PET) scans, single photon emission computed tomography(SPECT) scans, digital radiography images, mammography images and Fluoroscopy images. X-ray imaging is a valuable tool for diagnosis, monitoring, and treatment of many medical conditions, and its non-invasive, widely available, low cost and fast nature makes it a popular choice for many medical professionals. The proposed approach presents an algorithm for enhancing X-ray images, improving their visual appearance and making their content more useful and meaningful. The results of the algorithm show that enhanced images have a more natural look and provide accurate details of the objects in the X-ray images. Overall, this algorithm can aid in the diagnostic process by providing clearer and more detailed images for medical professionals to interpret.

ملخص البحث :

Cryptography is the practice of securing communication and information through the use of mathematical algorithms. Block ciphers use the Substitution-Permutation and Feistel network structure to encrypt information, while stream ciphers use a onetime key. GOST is a symmetric key encryption algorithm, which means it uses the same key for both encryption and decryption. The strength of the encryption depends on the key size and the complexity of the algorithm, and a longer key size and more complex algorithm will generally provide stronger encryption. This study suggests a hybrid encryption method based on the GOST block cipher, the modified Salsa (20) stream cipher, and multiple chaotic key levels to provide adequate security with random hardness as high as the NIST standard tests and to change the key schedule as security operations. The simple key scheduling of the GOST algorithm is a weakness, which can sometimes be the weak spot of the related key cryptanalysis method. The proposed solution, on the other hand, uses a new hybrid algorithm based on a modified version of Salsa (20), GOST, and multiple chaotic keys to get the best combination and stronger security. Because the brute-force attack needs 2256 likely keys to break a key, which is inconvenient, it should not be used in this situation