What is Computer Vision? What are the real world applications of Computer Vision?

Computer Vision (CV) Make computers understand images and video. To describe the world what we see in one or more images and to reconstruct its properties such as shape, illumination and color distributions. To duplicate the abilities of human vision by electronically perceiving and understanding the image. Computer vision is a field that includes methods for acquiring, processing, analyzing, and understanding images and, in general, high-dimensional data from the real world in order to produce numerical or symbolic information, e.g., in the forms of decisions. Computer vision is also described as the enterprise of automating and integrating wide range of processes and representations for vision perception. Computer vision covers the core technology of automated image analysis which is used in many fields. Computer vision is concerned with the theory behind artificial systems that extract information from images. Image data can take many forms such as video sequences views from multiple cameras or multi-dimensional data from a medical scanner. Computer vision is concerned with modeling and replicating human vision using computer software and hardware. It combines knowledge in computer science, electrical engineering, mathematics, physiology, biology, and cognitive science. It needs knowledge from all these fields in order to understand and simulate the operation of the human vision system.

 Computer vision (image understanding) is a discipline that studies how to reconstruct, interpret and understand a 3D scene from its 2D images in terms of the properties of the structures present in the scene. Trucco and Verri Computing properties of the 3D world from one or more digital images. Stockman and Shapiro To make useful decisions about real physical objects and scenes based on sensed images. Ballard and Brown The construction of explicit, meaningful description of physical objects from images. Forsyth and Ponce Extracting descriptions of the world from pictures or sequences of pictures.

Vision
Vision is the process of discovering what is present in the world and where it is by looking.

Computer Vision
Computer Vision is the study of analysis of pictures and videos in order to achieve results similar to those as by men.

Real World Applications of Computer Vision

1) Optical Character Recognition: Reading handwritten postal codes on letters and automatic number plate recognition. 2) Machine Inspection: This involves rapid parts inspection for quality assurance using stereo vision with specialized illumination to measure tolerances on aircraft wings or auto body parts. 3) Retail : Object recognition for automated checkout lanes. 4) 3D Model Building: It is fully automated construction of 3D models from aerial photographs used in systems such as Bing Model. 5) Medical Imaging: It involves registering pre-operative and intra-operative imagery or performing long term studies of peoples brain morphology. 6) Automotive safety: Detecting unexpected obstacles such as pedestrian on the street. 7) Match More: Merging computer generated imagery (CGI) with live action footage by tracking feature points in the more video to estimate 3D camera motion and shape of the environment. 8) Surveillance: Monitoring for intruders and analyzing highway traffic. 9) Biometrics: It is used for automatic access authentication as well as forensic applications.

Explain in brief about challenges to be faced while designing a computer vision system. Following are the challenges to be faced while designing a computer vision system: Computer vision has no consistent and unified methodology and terminology. Due to lack of clearly formulated sub goals it is difficult to measure progress. There is a gap between the theory and practice of computer vision. Sometimes methods are borrowed without justification of their adequacy to compute vision. Few theoretical criteria exist to evaluate the performance of vision algorithms. It is hard to automatically set the optimal values of the parameters. Experimental results appear which are statistically insignificant and irreproducible. What are different components in mathematical model of imaging? A mathematical model is a representation of the key characteristics and attributes of an existing or potential system or application which presents information about the make-up of the system by breaking it down into its suitable components. The mathematical model of imaging has several different components. 1) An image function is the fundamental abstraction of an image. Different image functions may be used to represent same image, depending on which of its characteristics are important. E.g., a camera produces an image on black and white film which is usually thought of as a real-valued function. Most images are represented by functions of two spatial variables f(x) = f(x,y) , where f(x,y) is the brightness of the grey level of the image. An important part of the formation process is the conversion of image representation from a continuous function to a discrete function. 2) A geometrical model describes how three dimensions are projected into two dimensions. 3) A radio metrical model shows how the imaging geometry light sources and reflectance properties of objects affect the light measurement (brightness) at the sensor. 4) A spatial frequency model describes how spatial variations of the image may be characterized into transform domain. e.g., Fourier transform may be used to transform the intensity image into the domain of spatial frequency. 5) A color model describes how different spectral measurements are related to image colors. 6) A digitizing model describes process of obtaining M-Vector discrete valued image functions f(x), usually of 1, 2, 3 or 4 dimensions.

Explain following 2D and 3D image transformation. 1) Translation 2D Transformation 1) Translation :Object at P(x,y) is moved to another position P(x,y) 2) Scaling 3) Rotation

[ ]

2) Scaling :Uses to modify the size of an object P(x,y) will change to P(x,y)

[ ] [ ] [

] [ ]

If a scaling factor is less than 1, the object is reduced in that dimension. If a scaling factor is greater than 1, object is magnified in that dimension. If a scaling factor is negative, the object is reversed in that dimension. 3) Rotation :Rotates an image at an angle.

3D Transformations a) Translation :Position P(x,y,z) is change to P(x,y,z)

x = x + Tx y = y + Ty z = z + Tz b) Scaling :Change x= Sx .x y= Sy . y z= Sz . z c) Rotation :the object

1 1 1

size 1 1

Around X-axis 1 z 1

z 1 Around Y-axis

z 1

1 z 1 Around Z-axis

z 1

z 1

Explain perspective transform used in imaging geometry. Explain the terms world co-ordinates and image co-ordinates. a) Perspective transformation produces perspective by viewing 3D space from an ordinary eye point. b) To perform perspective transformation the eye point and the Centre of Focus must be set. The eye point is the position 3D space is viewed and can be thought as of the position of camera lens. The center of focus is not the focus on the film inside the camera but is the point being viewed from eye point. c) With the perspective transform a point in 3D graphics is projected onto a plane where the line connecting the point in 3D space and the eye point intersect with the plane of the projection, the plane passing through center of focus and perpendicular to line of sight. Let P(x,y,z) denote a scene point and P(x,y,z) denote its image.

World and image coordinates World coordinates can be defined in terms of physical coordinates system that refers to the pixel coordinates of the original data frame. The coordinates of image points in the camera reference Frame can be obtain form pixel coordinates. What are extrinsic and intrinsic camera parameters? Explain. Extrinsic Parameters:The parameters that define the location and orientation of the camera reference frame with respect to know world reference frame. Examples: a) 3D translation matrix (T) describes relative positions of the origins of the two reference frames. b) 3 x 3 Rotation matrix (R)-an orthogonal matrix (RT R = R RT = I) that brings the corresponding area of the frames onto each other. Intrinsic Parameters: The parameters necessary to the pixel coordinates of an image point with the corresponding coordinates in the camera reference frame. Examples: a) Focal length f. b) Location of image center in pixel coordinates. The effective pixel size in horizontal and vertical direction. c) If required radial distortion coefficient K.

Thanks for all contributors to machine vision, computer vision.