Image performance in underwater robots is one of the most challenging problems for autonomous underwater robotics due to light transmission in water. Although image restoration techniques can effectively remove a haze from a damaged image, they require multiple images from the same location making it difficult to use in real time. Considering the positive effects of in-depth learning strategies on other image processing problems such as coloring or finding objects, a deeper learning solution is proposed. The convolutional neural network is trained in image retrieval techniques to capture one...

Heart disease is one of the foremost critical human diseases within the world and affects human life very badly. In heart conditions, the guts are unable to push the specified amount of blood to other parts of the body. Accurate and on-time diagnosis of heart condition is vital for coronary failure prevention and treatment. The diagnosis of heart condition through traditional medical records has been considered as not reliable in many aspects. To classify healthy people and other people with heart conditions, noninvasive-based methods like machine learning are reliable and efficient. Within...