A Method to Find the Desired Object in the Image based on VGG16

Finding the desired object in the image is a field of computer vision that called object detection. Object detection can be a general detector that identifies all objects in the image or can be used only for detect a specific object. The first case is a complex task and needs the massive dataset but the second case that is executable with less data and can be useful for robotic and industrial applications. purpose of this paper is close to second case and introduces a network to determine the location of a desired object in the image. proposal network is a combination of a pre-trained network, additional convolutional layers and several fully connected layer. This network able to estimate thepixel coordinate of wanted object bounding box in image from cluttered environment. In this paper, VGG16 is selected as a pretrained network and we created a dataset containing 525 images from 10 objects that are cluttered in each image. the model was evaluated on dataset and The results was impressive

کلیدواژه ها:

Object Detection ، Deep Convolutional Network ، Computer Vision ، Pre-trained Network ، VGG16