Image retrieval sorting method based on multi-attribute features and computing equipment

Abstract

The invention provides an image retrieval sorting method based on multi-attribute features and computing equipment, relates to the field of computer vision, and solves the technical problem that in the prior art, design based on independent weights neglects the synergistic effect between features, and consequently the accuracy of a user in querying a target image is reduced. The method comprises the following steps: acquiring a color histogram of a first to-be-sorted image; vectorizing the color histogram to obtain a plurality of attribute aggregation vectors, and taking the attribute aggregation vectors as a plurality of markers in the first to-be-sorted image; generating a marker combination element based on a preset association attribute; expanding the attribute aggregation vector based on the marker combination element to obtain a plurality of attribute combinations; setting a first priority of the first to-be-sorted image according to the expanded attribute combination; if the acceptance value preset by the user is not higher than the preset threshold value, obtaining the second priority again; and arranging the first to-be-sorted images according to an arrangement sequence corresponding to the second priority. The method and device are used in the process that the user retrieves the target image.

Images