Data Acquisition is using of image single hand based on the location, property, hand side and no. of figure count for higher accuracy for gesture recognition in sign language system. For testing purposes collected images are included in the database.
Image preprocessing and segmentation, consist image acquisition, segmentation process and morphological filtering process. To separate object and the background the segmentation is carried out. For segmentation, cluster detection and edge detection algorithm are used. The segmented hand image represents certain features. These features are further used for gesture recognition. Preprocessing is ready on the stored database.
Feature Extraction is the process of decreasing data dimensionality by encoding related data in a compacted and removing less discriminative data.
Skin Detection is used to search for the human hands and reject other skin colored objects while sign Recognition is dimensionality removing technique based on extracting the wanted number of major components of the multidimensional data.
In gesture recognition system for skin area detection, cluster algorithm is used. For boundary filling and pixel detection task, edge detection algorithm and pixel detection algorithm was used respectively. For conversion of text speech, API was used.