CSE3510 Exam Answers 1.Describe the format, structure and use of the GIF files for static and animated image processing. GIF stands for Graphics interchange format. The graphics interchange format is a loos less 8 bit/256 colour protocol for on line transmission and interchange of raster graphic data in a way that is independent of the hardware used in their creation or display. Most GIF files are simple images with good colour and detail representation. GIF files are usually accessed by the internet because it is designed for internet transfers. GIF files have built in image compression. Most GIF files able to be read by any GIF reader. Developed and copyrighted by compuServe, who provide a royalty free licence for use of the GIF standard. GIF File Structure GIG Signature Screen Descriptor Global colour map Image descriptor Local colour map Raster data GIF Terminator GIG signature and terminator GIG signature identifies GIF files GIF87a GIF identifies the file as a GIF image stream. 87a represents a version number. GIF terminator ends the GIF file. 3B Hex or ; (semicolon) signal ends. GIF reader may wait for user to enter CR at the keyboard or a mouse click. Application determine what happens next. Screen Descriptor Screen width in pixels (bytes 1 2) Screen hight in pixels (bytes 3 4) >>Image descriptor – McccOppp M = 1 for globle color map CCC = bits of colour resolution Ppp = bits/pixel in image >>Background color (256) (byte 6) >>Zeros (byte 7) Global colour map >>Up to 256 color definitions, each with Red intensity Green intensity Blue intensity >>(255,255,255) is white >>(180,180,000) is medium yellow. (000,000,000) is black. >>Good for use in internet (CGE, VGA) Image descriptor >>Gif creates for multiple images for file >>Permits local colour palette for image or use global definitions >>Defines where image is on screen >>Good for internet menu buttons >>Good for thumbnail images >>Caters for old and new GIF formats. Local Colour map >>Same format as global colour map >>Included only if I = 1 in Image Desc. >>Applies only to current image >>Reverts back to global descriptor at the end on the image >>Able to change the bits/pixel for image. Raster data >>Images may be sequential or interlaced >>Sequential images are processed left to right, top to bottom >>Interlacing images are processed in four passes of the same raster data >>Interlacing allows a rough picture to be displayed with 1/8 of the data. >>The Internet user can exit the picture if it is interlaced. Use of GIF files for static image processing. We use GIF files as static image processors because of its ability to compress loss Lesley. GIF is a lossless compression strategy, based on LZW compression (Lempel Ziv Welch algorithm) there is however significant colour loss quantisation of 24 bit images to 8 bits. For full screen, 8 bit images of moderate complexity, 4:1 compression is the average. Since this is a lossless scheme, the compression ratio can not be increased by trading off quality. Use of GIF files for animated image processing. GIF files permit extension blocks to simulate simple animation (e.g. Goldfish) Since a GIF data stream can contain multiple images, it is possible to process and display of such a file. Because the image descriptor allows for placement of the image within the logical screen, it is possible to define a sequence of images that may each be a practical screen, but in total fill the entire screen. There are some guidelines for handling the multiple image situations; they are: 1.there is no pause between images. Each is processed immediately as seen by the decoder. 2.Each image explicitly overwrites any image already on the screen inside of its window. The only screen clears are at the beginning and end of the GIF image process. Design of a system for multimedia conference proceedings Currently multimedia technology is used to incorporate graphics, animations, audio & motion video. >>>Multimedia hardware/software designed for conference proceedings provide all features of multimedia H/S in an integrated scheme for user identification with the system. >>>Primarily designed to overcome the shortcomings of traditional means of presentation. >>>Delivers, text, graphics, sound as an integrated whole providing extensive provision of random access to hyper media. >>>Typical conference proceeding is contained with desired papers of a conference in hyper text format interchanged with various talks of speakers in a movie format. >>>Hypermedia links provide random access to various parts of the system. >>>Includes a navigation shell, which provides hierarchical navigation. >>>Consists of transparencies including marks that were added during the talk. >>>Consists of hyper links containing relevant parts of proceedings &bibliography. >>>Users can also extend the system to create their own hyper links among objects that they seem relevant & keep notes of documents. >>>Users can get a brief overview of the talk by using the popup menu containing section titles of the talk or by skimming through the transparent >>>If the speaker for example mention a theorem without proof, users can jump to the hyper text to read the proof in detail. >>>Users can also do a search on keywords to find out who else is mentioned this keyword during the conference. >>>From there they can jump directly to the video of the second speaker of the relevant section, allowing the user to compare options of the relevant speakers. >>>Users can make various remarks on the subject being read & send these remarks to others by emailing them the file. >>>Other useful features are the option to jump directly to the beginning of a subsection within the talk using Jump to highlight button. >>>Users then presented with an ordered popup list of highlights of the talk from which the user can directly jump to. >>>Also it provides tools in a palette format that provide help, a map of the hypertext contents, bookmark facilities & a notebook. (So you know where you are now) >>>Users can also gain a history of notes they ve visited in a session. >>>Map facilities are provided so that the user can gain a geographical overview of that whole hypertext document. S1.Concepts and methodologies of optical character recognition and its use with pen based systems. Optical character recognition is the process of identification/recognition of printed or hand written text by a computer and representation in machine ASCII format for processing and manipulating. >>Involves photo scanning of text, character by character, analysis of image & translation into character code. >> Scanned bitmap image is analysed for light a dark areas in order to identify letters or numeric digits. >>Special circuit boards & computer chips designed for OCR. >>Typically involves three processes, Document analysis, Character Recognition & contextual processing. 1. Document analysis – requires reliable segmentation of scanned characters. For that requires good quality scanned document. >>Alignment used for this process >>Uses space recognition between characters. 2. Character recognition – use algorithm in which Feature Extractor & classifier is used. >>character image features are derived by feature extractor. >>matrix modelling used in which individual pixels are used as features. >>Comparison is done by comparing character with prototype template. >>Each comparison result s similarity measure between input character & template. >>Often uses Computer intelligence where computer looks general features such as open arias, closed shapes, diagonal lines. 3. Contextual processing – After recognition of character strings are then compared with dictionary. >>OCR output is stored in proprietary file format specific to OCR software. software highlights non recognisable characters. Post – OCR process converts file into one or more output formats in ASCII such as word, Adobe s, PDF. Pen based systems >>Need to train computer for individual user s writing style >>Training takes 4 8 hours >>Computer conditioned to user >>Reasonable accurate 90% >>Interactive – system prompts when it can t recognized character. Recognition Techniques for pen based systems. >>Same as for printed materials >>Less Exact representation >>Mach against under templates>>More flexible matching techniques. >>Cater for different sizes and slants >>More complex OCR programs. SQ17. Problems associated with hand written characters Hand written characters are not constant. From user to user & written one user characters the way characters are being written are vary, OCR programs won t be able to identify those characters problem arises when recognising text that is embedded, users various patterns, irregular backgrounds & various dimensions. In addition, handwritten characters can be merged together. In that case, it is really hard to identify individually. Different language combinations can also be used in the scanned paragraph which OCR cannot understand. SQ8.Gray Scale A greyscale images are simply images, which are only uses shaded of grey as colours. The reason for differencing such images from any other sort of colour image is that less information needs to be provided for each pixel. In fact a grey colour is one in which the red, green & blue components all have equal interesting in RGB Space and so it is only sensory to specify a style intensity values for each pixel, as opposed to three intensities needed to specify each pixel in a full colour image. Often, the greyscale intensity is stored as an 8 bit integer giving 246 possible different shades of grey from black to white. If the levels are evenly spaces than the difference between successive gearlevers is significantly better than the grey level resolving power of the human eye shades of grey are simulated by screens & printers by grouping bit patterns. Eg different grey levels will be provided by replication of preselected fillings of 3 x 3 box different bit patterns result in different shades of grey. Gary scale images are very common, in part because much of today s display and image capture equipments can only support 8 bit images. In addition, grey scale images are entirely sufficient by many tasks & so there is no need to use more complicated & hard to process colour images. S 18 Facilities provided by image editors >>Image editors also known as paint applications software in graphic >>Allow creating bit map images using the help of verity of painting tools. >>Include features of photo enhancement & retouching. >>Most common type of image format used is bitmap, because it is made with series of dots on grid. >>INSERTING – other images, text, graphics can add to the main image by inserting. >>Also allows creating drawings using lines & calls vector isolations. Vector illustration can be easily edited by moving points and changing colours of various objects. Can provide facilities of import of other images to gain greater quality of image. >>Provide facilities for editing points of image & re shaping structure. >>Allows changing resolution of images & graphics of image. >>Allow manipulating brightness & contrast of image >>Allow doing conversion into other image display formats such as greyscale or black & white. >>Allows facilities of creating dock able palettes in which you can check commonly used pallets in a big master pallet. >>Allow creation of layer styles. >>Provide facilities for various views of image by zooming in / out or get full screen. >>Allows tone adjustment of image in which you can adjust tone in terms of highlight, mid tone or shadow. >>Allows editing of various areas in image by creating highlighted variable areas & creation of new text images. >>Allow normal file menu facilities such as new, open, close save, ext… EDIT Menu Items à undo/cut/copy/past/clear/clipboard/select/duplicate/fill/flip/rotate/ crop/ Image Menuà tone adjustment brightness and contrast, Hue & Saturation/Invert/level adjustment. Convert Menu à Black and white/Greyscale. S 14. Describe a system fir indexing images in Microsoft access, using examples to illustrate the operation of the indexing system. Most image indexing involves identifying image with a descriptive title and/or comment field. All the keywords in the title or comment field can be used to index the images. MS Access table can be used to store the keywords & the image details. (image code or the path) eg. Sally and Anne going to school. Keyword sally, Anne School Image 1 The keyword of sally, Anne & School is being indexed as the above table & each keyword is an index pointing to the image. Doing a keyword search can retrieve images. Eg. If the user do a search by the name Sally it will run a SQL query & point to image 1. Image can be retrieved by writing the necessary code. Difficulties associated with indexing images. Indexing images & other textual information subsequently used in a content – based image retrieval system. Images will be identified by a descriptor title or a comment field, there are some difficulties associated with creating an image index. In designing such a image database system, two major issues are, 1. the amount and type of processing required when inserting new pictures into the database. 2. Efficient retrieval schemes by query processing. >>Searching captions (title) or keywords & names will not necessarily yield the correct information, since objects mentioned in the caption are not always in the picture and vice versa. >>performing a visual search for object of interest ( faces, finger prints) at query. This process is time consuming and can be expensive. >>in addition, the selective processing of text & picture at data entry time is clearly required. This could be time consuming process. >>eg. Extra care must be taken to ensure that the keywords are not being repeated when creating an associated caption for image. Otherwise, keyword search may return unrelated images. SQ 13. Three Small Images used to access large gif files Code to click filled circle Centre of the circle = (Cx, Cy) Radius R Mouse position = (Mx, My) D = distance between mouse and the center D = Sqrt (Mx Cx)**2 + (My Cy)**2); If D < R + 0.1 then In Else Out; Q4: Adv & disAdv of Tiff over BMP Adv >>Suitability for wide range of applications, its independence of a computers architecture, operating system and graphics hardware >>It handles black & white, greyscale & colour images well, allowing a user to adjust for the unique characteristics of a scanner, monitor or printer >>>Smaller file size compare to bitmap Disadv – >>>Requires significant programming effort to fully decode >>>Tiff data can be compressed in several different arrays >>>To be robust or fully functional, a tiff reader (program code that read tiff files) must support these various compression schemes >>>Complex file structure Q5: adv & disadv of GIF over Bmp Adv >>>Wide range of input/output device support GIF provides enough information & organise it well enough that widely differing input & output devices can easily exchange images >>>Support numerous platforms compuserve distribute the format specification freely>>>Support 8 bit coulour implemented as a palette of up to 256 colours and image sizes up to 64k x 64k pixels>>>Wide range of features LZW compression, sequencing or overlaying of multiple images, interlanced screen painting & text overlays>>>Smaller file size compared to bitmap Q7: fractal transforms >>>With the advance of the information age the need for mass information storage and retrieval grows. >>>Fractal transforms is a method to code any kind of image into mathematics equations. >>>These equations regenerate images to any size, larger or smaller than the original, which can be specified by the user >>>With fractal compression, the repetitive patterns of an image are identified and matched >>Two patterns are considered the same as long as one can be stretched, squeezed or rotated to like on another >>>The process continues until the image is the size specified by the user >>>This process is intensive, with compression taking longer than decompression >>>But compression rate of 100 to 1 can be achieved with little or no noticeable loss of image quality >>>An example of a photocopying machine that reduces the image to be copied and reproduces it three times of the copy, the output will then feed back as the input to the machine. After several iterations, the copies seem to be converging to the same final image, which is the attractor The initial image placed on the copier doesn t affect the final attractor, in fact it is only the position and the circulation of the copies that determines what the final image will look like >>>Since it s the way the input image is transformed that determines the final results of running the copier in a feedback loop, we only able to describe these transformation s>>>In real practice, choosing transformations of the form is sufficient to yield a rich and interesting set of attractors >>>Such transformations are called affined transformation. Whereby the image can be skew, stretch, rotate, scale and translate an input image >>>These transformations reduce the square to half its original size and copy it at 3 different locations in the same orientation >>>Each image is than formed from transformed copies of itself are call fractals which must have the detail of every scale storing images as collections of transformations could lead to image compression (eg fern leaf compression) Q8: Describe the concept of grey scale and its application to dot matrix and laser printers. >>>Grey scale is a colour mode made up of 256 shades of grey >>>These 256 colours includes black, white and 254 shades of grey in between them >>>Images in grey scale mode have 8 bits of information (2 to the power of 8th) >>>Shades of grey are simulated on screens and printers by grouping bit patterns >>>Dot matrix and laser printers does not produce grey dots but simulate different shades of varying the dot size so as to control the fraction of paper covered by ink >>>Each individual dot can have a different shade of grey >>>Halftone in printing is a continuous tone image, such as photograph, that has been converted into a black and white image >>>Halftones are created through a process called dithering, in which the density and pattern of black and white dots are varied to simulate different shades of grey >>>In order to simulate variable sized halftone dots in computer printers, dithering is used, which creates clusters of dots in a halftone cell. The more dots printed in the cell, the darker the grey. As the screen frequency gets higher (more cells per inch), there is less room for dots in the cell, reducing the number of shades of grey or colour that can be generated. >>>In low resolution printers, there is always a compromise between printer resolution (dpi) and screen frequency (lpi), which is the number of rows of halftone cells per inch. For example, in a 300 dpi printer, the 8x8 halftone cell required to create 64 shades of greys results in a very coarse 38 lines per inch of screen frequency (300 dpi divided by 8). However, a high resolution, 2400 dpi image setter can easily handle 256 shades of grey at 150 lpi (2,400 / 16). Q9: imaging facilities provided in MS access >>>Able to get images from many different ways >>>Getting images from clipboard, scanned images, copies from MS word etc >>>Can reference an image file directory >>>Able to link an image directly to the file without embedding it in MS access >>>Able to select text file or spreadsheet >>>Can insert object for OLE >>>Can click on the right mouse button to store images or add an object >>>Use image properties to change the image settings >>>Able to write events on the images Q: Describe the concepts and methodologies of OCR and its used with pen based computers >>>OCR involves reading text from paper and translating the images into a form that the computer can manipulate (for example, into ASCII codes). >>>All OCR systems include an optical scanner for reading text, and sophisticated software for analysing images. Most OCR systems use a combination of hardware and software to recognize characters or fonts >>>Scanned bitmap image is analysed for light a dark areas in order to identify letters or numeric digits. >>>Typically involves three processes, Document analysis, Character Recognition & contextual processing. 1. Document analysis – requires reliable segmentation of scanned characters. For that requires good quality scanned document. 2. Character recognition – use algorithm in which Feature Extractor & classifier is used 3. Contextual processing – After recognition of character strings are then compared with dictionary >>>Using a stylus or a pen based devices rather than a keyboard for input. This means that they also incorporated handwriting recognition features. >>>Handwriting recognition is a technique by which a computer system can recognize characters and other symbols written by hand. >>>Digitising tablet is an input device that enables you to enter drawings and sketches into a computer. It consists of an electronic tablet and a pen >>>The pen looks like a simple ballpoint pen but uses an electronic head instead of ink. The tablet contains electronics that enable it to detect movement of the cursor or pen and translate the movements into digital signals that it sends to the computer. >>>For digitising tablets, each point on the tablet represents a point on the display screen in a fixed manner. Pen based systems Need to train computer for individual user s writing style Training takes 4 8 hours Computer conditioned to user Reasonable accurate 90% Interactive – system prompts when it can t recognized character. Recognition Techniques for pen based systems. Same as for printed materials Less Exact representation Mach against under templates More flexible matching techniques. Cater for different sizes and slants More complex OCR programs. Q10: describe the use of hypertext on the Internet >>>Hypertext is a retrieval system that indexes words and creates threads between words in a text so that a highlighted word leads directly to another word or page >>>Is useful in conjunction with massive databases because it takes information retrieval to a new level, making information simpler to retrieve and find>>>To create a hypertext system, the contents of the database must be coded to link one subject with another, it than allows the user to jump from one topic to another, based on their immediate informational needs >>>For example, by selecting highlighted or usually underlined words or phrases, users can access pop up windows tat contain the detailed explanation of the text. >>>In multimedia applications they can access related audio messages and video sequences, without having to leave the main text of a document. >>>Hypertext links between ideas, text, documents, files, databases and audio/visual applications have to be created manually. Even with programming experience, this can be a very time consuming process >>>Another drawback of hypertext is that it is easy for any users to get lost or become frustrated because there appears to be no beginning or end to the database. Some system however solves this problem by recording the user s last position, which allows the user to return to the point where they left. Q20: correspondence system >>>Letters to council from residents >>>Scanned into system when they arrive & classified as new documents >>>Co ordinator looks at each document & directs it to person or area for processing >>>Problem is investigated and the letter is sent Db requirement >>>ID number for correspondence >>>Name & add of the sender >>>Data arrived >>>Current status >>>Department handling >>>Date of return letter >>>Reference to return letter Hardware requirements >>>Scanner >>>Scanning station >>>Display station >>>Print server >>>Printer >>>Fax serve r>>>Data server (database) Q24: describe imaging aspects of multi media systems >>>The use of computers to present text, graphics, video, animation, and sound in an integrated way >>>A CD Rom contains all the information for installing a product for example, will have a combination of text, audio and full motion video. >>>This allows the user to listen to warnings or music, view demonstration, read relevant documentation can be accessed from the same program >>>Multimedia conferencing systems consists of video and PC monitors, remote controls, speakers, software, cameras and microphones allows computers to pass data between two or more geographically separated locations, (eg. Multimedia Conference) >>>Some system offers a real time drawing capability which permits users to diagram a difficult concepts in a discussion forum or conference >>>From a desktop the user can select images to be enlarged, grouped, mailed, printed and copied for reference >>>Video mail concept, allows the users to incorporate video messages with data, graphics, and images. This media mail can be sent in encoded form to a dedicated messaging server for distribution to a speaker equipped PC. The message can be stored in the hard disk for playback or be downloaded to a CD for a manual distribution >>>Processing, storage, communication and high resolution display capabilities are required in any multimedia information system, therefore it needs high speed network interconnection among user workstations and geographically dispersed databases are required to manipulate and retrieve the unique data types. Q25: (a)major problems with transfer images >>>Primary problem of transferring images over a network is due to the typical file size of the image >>>Image files are of large size because of the large number of locations of being able to be accessed (pixels )>>>Since networks transmission medium such as twisted pair & coaxial cable have limited data transmission rates, transferring data associated with images require greater time >>>This is particularly a problem in the situation of a video conference in which the video image consist of about 30 frames per second & each frame is about 1mb in size >>>The limitations of such network transfers is due to the bandwidth of the transmission medium being a scarce resource >>>Furthermore transmission impairments such as attenuation cause problems in data transfer >>>Increase in data rate has potential effect of rising in cost of the network system & therefore its not justifiable (B) methods to overcome >>>The use of compression algorithms to images to achieve reduction in file size >>>Main concept is to reduce data in to an abbreviated form that still retaining the basic information contained in the image file >>>Various hardware & software devices developed to achieve compression of images file on transmission & decompression at receivers end >>>Another method is the sending of information in a hierarchical manner in which abbreviated information is presented first & provides user to view information which consist of larger volume of data later (eg. Displaying a thumbnail than able to display full image when click) >>>Various compression algorithms include JPEG, GIF, MPEG >>>Result is a reduction in file size