openCVIntro.pptVIP

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Motivation TM model accuracy is below 30% With this accuracy, the system cannot detect the overall concept of the image ? we cannot rely on the result! To remove the noisy words in an image We make use of Knowledge-base WordNet is a mirror of Human experience WordNet “Car” in the WordNet Car: a motor vehicle with four wheels; usually propelled by an internal combustion engine .self-propelled vehicle . wheeled vehicle . vehicle .instrumentation . artifact .object .entity Semantic Similarity Conceptual distance between two words Snow bear is more similar than snow beach Semantic similarity reflects spatial and temporal relations Semantic Similarity(2) Annotation results in images Fish, building, sea, coral, flower Sky, mountain, city Noisy keywords Semantic Similarity(3) Semantic Grouping fish sea coral building flower choose Relevant keywords remove Detect Concept Results Results(2) Results(3) Yohan Jin, Latifur Khan, Lei Wang, and Mamoun Awad, Image Annotations By Combining Multiple Evidence WordNet, In Proc. of ?13th Annual ACM International Conference on Multimedia (MM 2005), Singapore, November 2005 (2005), 715. Contents OpenCV Installation “Hello World” Program of OpenCV OpenCV Library Inside Image Annotation Image Classification with OpenCV Video Processing with OpenCV Image Classification with openCV Image Classification Framework Mamoun Awad, Yohan Jin, Latiur Khan, George Chen, and Fehmi Chebil, A Framework for Image Classification the IEEE 2006 Southwest Symposium on Image Analysis and Interpretation openCV used for low-level feature extraction Low level feature extraction based on Grid of images Texture : Wavelet transform ( LL,LH,HL bands coefficients-3dimension) Color: average of RGB,LUV color features (6 dimension) Color-Histogram :(7*7) of H and S (49 dimension) Ground Truth Training Query by Example Contents OpenCV Instal