Bilgisayar görüşü ve derin öğrenme içinotomatik algılama ve lokalizasyon

Machine learning techniques such as neural networks, which include concepts that have been found in organic nerve systems, are becoming more popular. A biological neuron is a cell that consists of a body, many smaller protrusions called dendrites, and a single long protrusion called an axon. Axons are the long protrusions that connect the dendrites to the body of the neuron. On the surface of the dendrites, there are several sites known as synapses, which are where axons from other neurons are joined to the neuron in question. An electric impulse is fired by the axons every now and then, and this changes the permeability of the cell membrane, causing a modest rise in the voltage within the neuron body. The greater the number of activations that come from other neurons, the greater the increase in voltage. A center is located at the base of the axon, and it is responsible for permanently measuring the voltage. The goal of semantic segmentation is to identify pixels belonging to the same object or class in the image and cluster them together. This method is valuable in multiple applications and projects such as detecting brain tumors, creating semantic maps with civilians for autonomous vehicles, face segmentation viii or land usage classification. we used semantic segmentation, Semantic segmentation provides pixel-wise boundaries around screens, and as screens do not entirely overlap, it yields similar results as instance segmentation with significantly less complexity.