Image recognition search method and related device

Abstract

The invention discloses an image recognition search method and a related device, and the method comprises the steps of firstly, judging whether the intention type of an input image is a commodity type or a non-commodity type through an image intention classification model; then, when the intention category of the input image is a commodity category, extracting image features of the input image through an input image recognition model; and finally, searching at least one commodity image similar to the input image in the commodity image database according to the image features of the input image, and forming a similar image set of the input image. Therefore, the image intention classification model is used for performing intention category judgment on the input image, and the operation of searching the commodity by the image is performed only when the judgment result is the commodity category, so that the purpose of filtering the input image of the non-commodity category is achieved, namely, the operation of searching the commodity by the image is not performed on the input image of the non-commodity category; therefore, waste of computing resources is avoided, and the computing resources are greatly saved.

Description

technical field [0001] The present application relates to the technical field of image processing, in particular to an image recognition search method and a related device. Background technique [0002] With the rapid development of image recognition technology, in daily life, you can easily take pictures of clothes, cosmetics, daily necessities, etc. you are interested in, and then use the images to search for products. Links to the same product or similar products, so as to meet the needs of users to quickly search for products of interest. [0003] At present, the specific implementation process of searching for products by images is: firstly use the deep learning model to extract the image features of the input image, and then calculate the distance between the image features and the image features of each product image in the product image database, so as to return images similar to the input image. image of the product to get links to the same product or similar produ...

AI Technical Summary

Problems solved by technology

[0004] However, the inventor has found through research that the categories of input images are various, and may be commodity images, and may also be non-commodity images, and non-commodity images do not actually need to search for commodities by images; while the prior art In this method, the above operation of searching products by images is performed on the input images, that is, the operation of searching products by images is also performed on non-commodity images, which is extremely easy to cause a waste of computing resources

Images