International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056 Volume: 04 Issue: 04 | Apr -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 3480 RECENT TRENDS IN 2D TO 3D IMAGE CONVERSION: Algorithms at a glance Avi Kanchan 1 , Tanya Mathur 2 1 Student, Department of Computer Science Engineering, KIET, Ghaziabad, India 2 Asst. Professor, Department of Computer Science Engineering, KIET, Ghaziabad, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - We present a paper on state of the art methods of 2D to 3D image conversion. In this modern era, 3D contents are dominated by its 2D counterpart. Today there exists an urgent need to convert the existing 2D content to 3D. Mainly, these conversion methods are categorised in an automatic method and semi-automatic method. In an automatic method, human intervention is not involved, whereas in semi- automatic method human operator is involved. The main difference between 2D and 3D images is clearly the presence of depth in 3D images which makes the calculation of depth the most important factor. Until now many researchers have proposed different methods to close this gap. This paper describes and analyses algorithm that uses monocular depth cues and by learning depth from examples, establishing an overview and evaluating its relative position in the field of conversion algorithms. This may, therefore, contribute to the development of novel depth cues and help to build better algorithms using combined depth cues. 1. INTRODUCTION We must begin our journey by taking issue with the philosophical adage that "a picture is worth a thousand words." It is my belief that a picture cannot begin to convey the depth of human experience and wisdom embedded in the words of Shakespeare. Nonetheless, pictures do contain a wealth of information and have been used throughout the centuries as an important and useful means of communication. An image is a picture representing visual information. A 2D image has only two dimensions height and width, while in a 3D image along with height and width it contains a third parameter called depth. The 3D image provides more information and gives better real-time world experience than 2D image. The advent of innovative 3D technology and accruing sales of 3D consumer electronics have accompanied an increase in demands of more and more 3D technology. Despite the advent of 3D technology, the availability of 3D content is still hindered by that of its 2D correspondent. To close this gap, many 2D-to-3D image conversion methods have been proposed. Two approaches to 2D to 3D conversion can be loosely defined: semi-automatic conversion and automatic conversion. In semi-automatic conversion, a skilled operator assigns depth to various parts of an image or video. In automatic methods, operator interference is not required despite a computer algorithm automatically measures the depth for a single image. Automatic methods estimate shape from shading, structure from motion or depth from defocus. , have not yet achieved the same level of quality for they rely on assumptions that are often violated in practice. Methods involving human operators have been most successful but also time- consuming and costly. The main difference between 2D and 3D images is clearly the presence of depth in 3D images which makes the calculation of depth the most important factor in the conversion of images from 2D to 3D. There are two steps in 2D to 3D conversion process: depth estimation for a given 2D image and depth-based rendering of a new image in order to form a stereo pair. While the rendering step is well understood and algorithms exist that produce good quality images, the main problem is in estimating depth from a single image. Several methods have been proposed for the same. Out of these, we shall study mainly two methods. First, calculating the depth using monocular depth cues and then by learning depth via a simplified algorithm that learns the scene depth from a large database which is having an image and depth pairs. To compare these two methods, we use the generation of a depth map. A depth map is a 2D function that gives the depth (with respect to the viewpoint) of an object point as a function of the image coordinates. The depth map is a kind of image which is composed of the gray pixels defined by 0 ~ 255 values. The "0" value of gray pixels stand for that "3D" pixels are located at the most distant place in the 3D scene while the "255" value of gray pixels stand for that "3D" pixels are located at the most near the place. In-depth map, each depth pixel would define the position in Z-axis where its corresponding 2D pixel will be located. It is called as pixel-by-pixel which produces a reasonably good 3D image, it is now widely used for producing 3D contents, especially the multi-view 3D contents for 3D digital signage. Figure 1: A 2D image and its depth map 2. ESTIMATING DEPTH BY LEARNING DEPTH FROM EXAMPLES The proposed method is an automatic conversion for images. Mining techniques based on image parsing have been used