System and method for searching image of goods

A system for searching one or more images of goods, the system comprising:

• a shopping mall server communicationally connected to a plurality of user terminals, the shopping mall server configured to provide at least one of the user terminals with similar goods images matching a plurality of input goods images transmitted from the at least one of the user terminals as a search result for recommended goods;
• and
• an apparatus communicationally connected with the shopping mall server, the apparatus having a processor configured to receive information on the input goods images from the shopping mall server, calculate feature vectors corresponding to the input goods images, and transmit the calculated feature vectors to the shopping mall server,
• wherein the shopping mall server comprises a first database configured to store held goods images for respective goods held by the shopping mall server and feature vectors corresponding to the held goods images, and a second database configured to store information on the input goods images transmitted from the at least one of the user terminals, and
• wherein the processor of the apparatus communicationally connected with the shopping mall server is configured to:
• receive first and second input goods images to be input as metadata;
• convert information on the received first and second input goods images into vector image data, and execute a deep learning algorithm to analyze the information on the first and second input goods images;
• extract feature vectors for each of the first and second input goods images to which the deep learning algorithm is applied;
• and
• generate a first input image feature vector and a second input image feature vector based on the extracted feature vectors to provide the feature vectors corresponding to the input goods images to the shopping mall server.