Method for determining point of interest for user, electronic device and storage medium

An electronic device, comprising:

• a processor;
• a memory for storing the processor-executable instructions;
• wherein the processor is configured to read the executable instructions from the memory and execute the instructions to implement the following steps of:
• in response to receiving a point of interest query instruction of a user, determining point of interest classification information about points of interest to be queried based on the point of interest query instruction, wherein the point of interest classification information comprises information about whether to obtain a type to which the point of interest to be queried belongs, and a classification type in a case of obtaining the type to which the point of interest to be queried belongs;
• acquiring body state information about a user, vehicle state information about a vehicle carrying the user, and surrounding environment information about a pre-set range around the vehicle, wherein the body state information comprises at least one of an orientation, a line-of-sight direction, or a gesture pointing direction;
• wherein the surrounding environment information comprises points of interest on both sides of a road;
• and wherein the vehicle state information is obtained based on sensors on the vehicle, and the body state information is obtained based on a camera in the vehicle;
• determining a set of candidate points of interest based on the vehicle state information, the surrounding environment information and the point of interest classification information;
• determining scores respectively corresponding to each of the candidate points of interest in the set of candidate points of interest based on the body state information about the user, vehicle state information and a pre-set scoring rule, wherein the pre-set scoring rule comprises at least one of a score predictive model with a neural network structure or a scoring mapping function taking the body state information and the vehicle state information as parameters;
• and
• determining a target point of interest corresponding to the point of interest query instruction based on scores respectively corresponding to each of the candidate points of interest;
• wherein the determining scores respectively corresponding to each of the candidate points of interest in the set of candidate points of interest based on the body state information about the user, vehicle state information and a pre-set scoring rule comprises:
• determining a query start point, a query vector in a map coordinate system and a pitch angle relative to the map coordinate system of an intended direction of the user based on position information of the vehicle and the body state information about the user, wherein the position information of the vehicle is determined based on the vehicle state information, wherein the pitch angle relative to the map coordinate system is an included angle between a direction vector of the intended direction of the user and the map coordinate system;
• determining a target included angle area range based on the query start point, the query vector and a pre-set included angle threshold;
• determining a set of target candidate points of interest within the target included angle area range from the set of candidate points of interest;
• determining a distance respectively between each of the target candidate points of interest and the query start point based on coordinate information about the query start point and each of the target candidate points of interest in the set of target candidate points of interest;
• and
• determining scores respectively corresponding to each of the target candidate points of interest based on road information, a point of interest query type, coordinate information respectively corresponding to each of the target candidate points of interest, a distance respectively between each of the target candidate points of interest and the query start point, and the pitch angle, wherein the point of interest query type comprises one or more point of interest types, and is determined according to the point of interest classification information.