The present application discloses a method,
system and
server for locating products in a supermarket. The method comprises: electronic shelf labels transmitting signals for measuring angles of arrival and signals for locating neighbor shelf labels, receiving locating neighbor signals of other shelf labels, and determining neighbor information; base stations measuring angles of arrival of shelf
label signals; a cloud positioning
server determining preliminary positions of the shelf labels on the basis of the angles; determining a neighbor relationship of all the shelf labels on the basis of the neighbor information; on the basis of the neighbor relationship, screening a plurality of preliminary locating positions of each
electronic shelf label to obtain an optimized locating position of each
electronic shelf label; a map
server predicting a preliminary product display position aggregation result of each
electronic shelf label on the basis of the optimized position, and mapping the preliminary product display position aggregation result onto a supermarket map image; and optimizing each preliminary aggregation result in the image on the basis of the following factors to obtain a result representing a locating result of each product bound to the corresponding shelf
label: the neighbor relationship, an actual distance, the category of a display position, and the ratio of the area of the display position to the number of shelf labels having been aggregated.