Multi-layered Object Identification and Detection Using Deep CNN Detector

There are many people today engaged in several activities spread over numerous locations. For a safe environment and safety, it may be necessary to describe the types of people and their behaviour at certain public meetings. Manpower can be quite expensive and not very useful when trying to observe a crowd. The existing technology makes it both more affordable and effective. We can suggest deploying closed-circuit cameras to keep an eye on a crowd of individuals and their behaviour. In this research, we have proposed the algorithm for identifying object by using deep Convolutional Neural Network (CNN) detector. This algorithm is used to extract the features from local datasets. Convolution, pooling, and affine layers are frequently combined in CNN. Convolution layers are created using a number of separate filters, each of which glides over the image to generate a map. The pooling layer then works to condense the visual representation. This proposed algorithm is more accurate and needs less processing time for evaluating the quality of object.