Content Based Video Retrieval API

CBVR API
siat.vdpl.videoRetrieval
Method Params Returns
index(VideoDataset) VideoDataset: Collection of videos to index Count of processed videos
retrieve(QueryKeyword, Filter) QueryKeyword: Text Keyword to retrieve videos against
Filter: Optional, Return criteria such as top-k, date, length etc., default value is NULL
Matched Video(s) or NULL
retrieve(QueryImage, Filter) QueryImage: Image to retrieve videos against
Filter: Optional, Return criteria such as top-k, date, length etc., default value is NULL
Matched Video(s) or NULL
retrieve(QueryVideoClip, Filter) QueryVideoClip: Video Clip to retrieve videos against
Filter: Optional, Return criteria such as top-k, date, length etc., default value is NULL
Matched Video(s) or NULL


siat.vdpl.videoRetrieval
Method Params Returns
extractSpatialFeatures(Frame, AlgoRef) Frame: Video frame to extract Spatial Features From
AlgoRef: Reference to the algorithm object to use for feature extraction
Feature Vector
extractTemporalFeatures(Frame[], AlgoRef) Frame: Video frame to extract Temporal Features From
AlgoRef: Reference to the algorithm object to use for feature extraction
Feature Vector
vlbp(Frame, preFramre, sucFramre) Frame: Target video frame to process
preFrame: Preceding video Frame to current Frame
sucFrame: Successor video frame to current Frame
Feature Vector
vltp(Frame, preFramre, sucFramre) Frame: Target video frame to process
preFrame: Preceding video Frame to current Frame
sucFrame: Successor video frame to current Frame
Feature Vector


siat.vdpl.videoRetrieval.Similarity
Method Params Returns
euclidean(srcFeatureVector, dstFeatureVector) dstFeatureVector: Source Feature Vector
dstFeatureVector: Destination Feature Vector
Similarity Measure between two Feature Vectors
cosine(srcFeatureVector, dstFeatureVector) dstFeatureVector: Source Feature Vector
dstFeatureVector: Destination Feature Vector
Similarity Measure between two Feature Vectors