Object detection from images using deep learning

The main goal of our thesis is to study deep learning and some of its different frameworks then we focus on object detection with deep learning technique and we compare which methods is suitable to dog's detection. For Object detection needs to process the classifying and locating objects in image. Object detection is a difficult issue that computer vision and machine learning involved. By applying various deep learning techniques, machine learning and computer vision achieved the most impressive results, for that deep learning become the most popular and widely used in research community. Convolutional neural network is a subtype in deep neural network that beat some contest in computer vision because is achieved good results and suited well in object detection tasks. Based on new examination of discoveries on new researches, the technique that is fit for our works is Faster Region Convolution Neural Network (Faster R-CNN) that we tend to used it to classify numerous varieties of dogs in image. Faster R-CNN for object detection and classifying beats several different ways particularly for the detection purpose. We chose faster R-CNN method because it's accuracy and speeds. We present a Faster R-CNN trained to classify and detect Dogs from image.