smart traveller with visual translator

26
Smart Traveller with Smart Traveller with Visual Translator Visual Translator

Upload: candice-cross

Post on 02-Jan-2016

34 views

Category:

Documents


2 download

DESCRIPTION

Smart Traveller with Visual Translator. What is Smart Traveller?. Mobile Device which is convenience for a traveller to carry E.g. Pocket PC, Mobile Phone. What is Visual Translator?. Recognize the foreign text and translate it into native language Detect the face and recognize it into name. - PowerPoint PPT Presentation

TRANSCRIPT

  • Smart Traveller with Visual Translator

  • What is Smart Traveller?Mobile Device which is convenience for a traveller to carryE.g. Pocket PC, Mobile Phone

  • What is Visual Translator?Recognize the foreign text and translate it into native languageDetect the face and recognize it into name

  • RequirementsSimple (Computational low power)Lightweight (Low Storage)User Friendly

  • Core Pattern Recognition Model

  • Character RecognitionLanguage: KoreanTarget: Sign, GuidepostContrast in ColorPrinted Character

  • Image SegmentationBinarizationUsing Color Histogram to binarize the image for the background and the characterText Region SegmentationUser Define MethodEdge Detection with horizontal and vertical projectionsStroke ExtractionLabeling of connected component Algorithm

  • Feature ExtractionStroke FeaturesNumber of Junctions, CornersAny HoleGabor Features

  • RecognitionMinimum Euclid DistanceLearn the Decision Tree by training examples

  • Demo

  • Face DetectionOutlineFind Face RegionFind the potential eye regionLocate the iris and eyelids

  • Find Face Region - Color-based modelWe used this method because of its simplicity and robustness.Usually RGB color model will be transformed to other color modes such as YUV (luminance-chrominance) and HSB (hue, saturation and brightness)

  • YUVWe use YUV or YCbCr color model.Y component is used to represent the intensity of the imageCb and Cr are used to represent the blue and red component respectively.

  • YCbCr ImageY, Cb ,Cr component imageY Cb Cr

  • Representation of skin colorWe just use a simple ellipse equation to model skin color.

    CbCr

  • Representation of skin colorThe white regions represent the skin color pixels

  • Color segmentationWe distribute some agent in the image uniformly.Then each agent will check whether the pixel is a skin-like pixel and not visited by the other agent.If yes, it will produce 4 more agents at its four neighboring points.If no, it will move to one of four neighboring points randomly and decrease its lifespan by 1. When its lifespan becomes zero, it will be removed from the image.

  • Color segmentationThis agent produce 4 more agents

  • Color segmentationThe advantage of this algorithm is that we need not to search the whole image.Therefore, it is fast.

  • Color segmentation 19270 of 102900 pixels is searched (about 18.7%)There are 37 regionsEach color regions represent each regions searched by a father agent

  • Eye detectionAfter the segmentation of face region, we have some parts which are not regarded as skin color.They are probably the region of eye and mouthWe only consider the red component of these regions because it usually includes the most information about faces.

  • Eye detectionWe extraction such regions.The red region represent the region which is not skin color.

  • Eye detectionWe do the following on the regions of potential eye regionHistogram equalizationThresholdTemplate matching

  • Eye detectionHistogram equalization

    Threshold with < 49

    Template Matching

  • Locating the iris and eyelidsWe plan to use the following methods to improve the face detectionWe can use these methods to locate the iris and eyelid precisely.

    Template matchingCorrelation variance filterDeformable template

  • END