Image retrieval method and device based on sparse representation

[0075] like figure 1 As shown, a flow chart of a sparse representation-based image retrieval method provided by the present invention, the method includes the following steps:

[0076] Step S1, input an image set, and perform preprocessing on the input images in the image set.

[0077] Step S2, using a group sparse feature selection strategy to select the feature information of the input image and the image database to form an image feature library.

[0078] Step S3, performing a specific metric comparison according to the features of the input image and the features in the image database, calculating the similarity, and obtaining a primary matching result.

[0079] Step S4, outputting an image similar to the input image according to the magnitude of the similarity.

[0080] like figure 2 As shown, it is a flowchart of step S1, and step S1 includes the following steps:

[0081] Step S11, input image set.

[0082] Step S12, performing size normalization on all input images i

[0102] like Image 6 Shown is a functional block diagram of an image retrieval device based on sparse representation provided by the present invention. The device includes: an input image preprocessing unit 1 , a feature selection unit 2 , a comparison calculation unit 3 and an output image unit 4 . The input image preprocessing unit 1 is configured to input an image set, and perform preprocessing on the input images in the image set. The feature selection unit 2 is configured to select the feature information of the input image and the image database by adopting a group sparse feature selection strategy to form an image feature library. The comparison calculation unit 3 is configured to perform a specific metric comparison according to the features of the input image and the features in the image database, calculate the similarity, and obtain a primary matching result. The output image unit 4 is configured to output an image similar to the input image according to the magnitud