Target detection method of full-fusion neural network based on semi-packet convolution

The invention discloses a target detection method of a full fusion neural network based on semi-packet convolution. The method comprises the following steps: establishing a semi-packet convolution module; building a full fusion neural network; acquiring a data set of a target object in the elevator car; dividing images in the data set randomly into a training data set and a test data set in proportion, labelling target objects and persons in the images of the training data set, generating label files, inputting all the images of the training data set and the label files into a full fusion neural network for training, obtaining and trained weight files; obtaining the positions and detection confidence of the detected target object and person in the image; and screening out credible targets, removing repeated target frames, and judging whether the targets enter the elevator car or not. According to the method, while the quality of feature extraction is ensured, the parameter quantity of convolution operation is