46results about How to "Optimize search results" patented technology

The invention provides a video index establishing method, comprising the following steps: A1, motion activity descriptors of all video frames which are contained in all video lenses are extracted, the motion activity descriptors which are corresponding to a plurality of video lenses and are obtained by extraction are trained, thus generating a corresponding rule set; B1, the rule set is queried according to the motion activity descriptors which are corresponding to the various video lenses, thus obtaining the corresponding type attributes of the lenses of the various video lenses; C1, a video index is established according to the type attributes of the lenses of the video lenses and the motion activity descriptors. The invention simultaneously provides a video index establishing device, an extraction method and a device of the motion activity descriptors, which can meet the delicate and the effective query requirements of a user and improve the user experience.

The invention provides a fast field tire trace pattern retrieval algorithm. Characteristics of a to-be-tested image is analyzed first, and the image is preprocessed; the characteristics that have the rotating, zoom and shift invariant performance, and satisfy the certain noise immunity and slight defect invariant performance of stock images and the to-be-tested image are selected within the margin of algorithm errors; finally the selected characteristics are combined to operate characteristic query. The fast field tire trace pattern retrieval algorithm considers both the whole and local characteristic query strategies, has the rotating, zoom and shift invariant performance, is insensitive to noise, has the obvious advantage in terms of classified query, and obtains a good query result when a test is operated based on a large sample data base. Particularly, the query result is better aimed at the field tire trace pattern retrieval algorithm which is lower in contrast ratio, larger in fragment, and distorted.

The invention belongs to the technical field of image retrieval, in particular to a structured image description method. The structured image description method comprises the steps that an image for training is obtained, and three-layer tree-shaped structure label is established for each object in the image, so that a training set is formed; the bottom-layer characteristics of each object of the image in the training set are extracted, all candidate classes, subclasses and classifiers with corresponding attributes are obtained through training, and therefore intermediate data required for modeling of the next step are formed; a conditional random field model is established and model parameters are obtained through training; image segmentation is firstly conducted, objects contained in an image to be described are segmented, and the bottom-layer characteristics of each object of the image to be described are further extracted; tree-shaped structure label of each object of the image to be described is predicated through the established CRF model and the model parameters obtained through training and according to the maximum product belief propagation algorithm. According to the structured image description method, the distinction degree between images can be improved and a good retrieval result is generated.