Cross-media Chinese herbal medicine plant image searching method based on deep learning

Abstract

The invention discloses a cross-media Chinese herbal medicine plant image searching method based on deep learning. The method comprises the following steps that 1, on the basis of OCR, text structural processing is performed, and plant classification description characters are extracted from books such as 'plant taxonomy'; 2, a Chinese word segmentation tool is used for performing preprocessing including word segmentation and stop word removal on all plant classification description characters; 3, a word2vec algorithm is used for generating word vectors according to a description text; 4, Fisher Vector is used for encoding the description text; 5, a convolutional neural network is used for performing training on an image set, and the network structure is made to converge to the best state; 6, the full last but one connection layer output of the convolutional neural network is extracted as an image feature vector; 7, textual features and image features are fused; 8, a linear kernel SVM classifier is used for performing model training; 9, during user retrieval, images can be input, plant texts can be described, and then a final image retrieval result is obtained through the steps 4, 6, 7 and 8.

Description

technical field [0001] The invention specifically relates to the field of information retrieval, in particular to a deep learning-based cross-media image retrieval method for Chinese herbal medicine plants. Background technique [0002] Chinese herbal medicine plant image retrieval technology is a part of differential image retrieval, which aims to obtain the category of the Chinese herbal medicine image retrieved by the user according to the picture and description text input by the user (at least one of the two). Accuracy problems encountered in image retrieval, adding corresponding plant description information such as leaf conditions can improve the accuracy of image retrieval. [0003] In the current common image retrieval, local feature description operators such as SIFT are usually used to generate local features, and then feature encoding methods such as Bag of Visual Words (BoV) are used for feature encoding, and then non-linear kernel function Support Vector Machin...

AI Technical Summary

Problems solved by technology

[0004] In actual operation, the nonlinear kernel function support vector machine (SVM) often takes a long time for data processing, and the performance degradation caused by time complexity is very obvious when facing a large data set

Images