Method, device, equipment and medium for detecting maturity and spatial position of tomato

The invention discloses a tomato maturity and spatial position detection method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a tomato image data set, marking the position and maturity type of a tomato in an image in the data set, and training a maturity detection network model through the tomato image data set and the marking; the maturity detection network model is improved based on a deep learning YOLOv4-tiny network model, Backbone is CSPdarknet53-tiny, and attention mechanisms are respectively added in the last two layers of output and an FPN (Fabry-Perot Network) structure of the Backbone; constructing a trained maturity detection network model and a detection model for obtaining the three-dimensional space position of the tomato through a computer vision technology; and inputting the image of the tomato to be detected into the detection model, and outputting the maturity and the spatial position of the tomato in real time. According to the method,