Carriage crowding degree detection method

The invention discloses a carriage congestion degree detection method, and relates to the technical field of target detection. Comprising the following steps: acquiring an image in a carriage, and dividing the image into an aisle area picture and a seat area picture; detecting and counting passengers in the seat area picture by using an improved target detection model YOLOv7, and detecting and counting passengers in the aisle area picture by using an improved density estimation model MCNN; adding the number of passengers in the seat area and the number of passengers in the aisle area to obtain the total number of passengers in the carriage, dividing the total number of passengers in the carriage by the total area to obtain the passenger floor area, and judging the carriage crowding degree according to the passenger floor area. According to the method, the target detection model YOLOv7 and the crowd density estimation model MCNN are improved, the two improved models are combined for use to count the total numb