Object detection and recognition is a sub filed of computer vision that is greatly based on machine learning. From the past few decades, it is observed that, the field of machine learning has been overwhelmed by deep neural networks called as CNN, i.e. Convolution Neural Networks. Computational power and the availability of the data are the two important characteristics to made CNN as powerful and such kind of neural network is well appropriate for image processing application for example object detection and recognition. The model is trained by identifying numerous features such as pixel positions, corners, hues, edges in the image and merge these features in to more complex shapes. To detect the objects, the model must compute the locations of the objects and classifies accordingly. There exists several methods to detect the objects using deep neural networks module in deep learning and works based on a pre-trained networks via Caffe, Tensor Flow and Torch/Py Torch for image classification. The work is comprised by combining Mobile Nets and Single Shot Detectors (SSDs) and more efficient and fast to detect the objects using deep learning. Our experimental results shows the better results compared with other methods for example Faster R-CNN, YOLO, and SSDs.