Description

The Video will be retrieved using content based video search for the input of speech and video text content. There are many lecture video data available online so need to detect text from that lecture video. There are many methods are available for text retrieval from video. Here a more efficient method has been proposed. In this first step video will be converted to frame. During frame conversion videos are separated into image format. Then preprocessed the each frame, in this step noise has been removed from each frame. After that canny edge detector has been applied to predict the edge. This process is repeated until all frames edge will be detected. Next feature has been extracted from each frame. In this process used HOG feature extraction method for extracting the feature. Finally SVM classifier is used to classify the text frame from the all detected frames. The Canny edge detector is an edge detection operator that uses a multi-stage algorithm to detect a wide range of edges in images. This edge detector method detects the edge in our frames. Histogram of Oriented Gradients (HOG) are feature descriptors which counts occurrences of gradient orientation in localized portions of an image. With HOG feature extraction method extracts gradient values of all frames. Finally SVM classifier is used for classification. Support vector machines are supervised learning models with associated learning algorithms, given a set of training examples, each marked as belonging to one of two categories, SVM training algorithm builds a model that assigns new examples into one category or the other, making it a non-probabilistic binary linear classifier.