Object detection and image segmentation are the procedure of object identification and organization. The model in this project is fabricated using YOLO Neural networks. In this project, we hope to achieve the same by training our CNN deep learning network on various datasets such as COCO acronym of Common Object in Context. The aim of this project is to disclose and classify objects which are fed to our network in different formats(png,jpeg, avi, mp4, etc.) in minimum time as well as maintaining a high accuracy. We see object identification even as backsliding trouble up to graphically bisect bounding containers also similar with class prospect. Unit neural chain forecasts enclose containers together with class chances immediately from filled pictures in single calculation. Since the entire identification duct or tube happen in a unit grid, it might be reformed everywhere quickly on identification representation. Our combine system is very rapid. Picture Segmentation happens in the technique of dividing a computerized picture within many parts. Segmentation, basically define as technique of set a flag to every singlepixelinsideapictureso thatpixelswhich has same flag have certain properties. And also, in this report we have given a short outline about some segmentation techniques used in picture processing like clustering, edge based, region based, model based, etc. Our simple YOLO model processes pictures at 45 fps. The small-scale category of the grid, Rapid YOLO, approach at 155 fps. Contrast to state-of-the-art recognition method, YOLO put together many localizations inaccurate still is smaller likely to forecast wrong positives on environment. This paper addresses the recognition, object detection and segmentation issues in whitebackgroundphotoswithdeeplearningmethod.Inparticular,wefirstlytrainarecognitionmodelbasedonYOLO to judge whether a photo is white background

Image Segmentation; Object Identification; Neural Networks; YOLO; COC.