Document text detection model training method and document text detection method

The invention provides a document text detection model training method and a document text detection method. The training method comprises the steps that training set images are acquired, and the training set images comprise receipt images and position labels corresponding to text areas in the receipt images; performing feature extraction on each document image to obtain a first feature map; performing feature fusion on the first feature map to obtain a second feature map; obtaining a probability graph and a threshold graph based on the second feature graph, and carrying out differentiable binarization processing on the probability graph and the threshold graph to obtain a binary graph and a text region position coordinate of the document image; and iteratively updating each parameter based on the position label, the position coordinate of the text region and a preset loss function, and when the preset loss function converges, obtaining a trained target document text detection model. The method can effectivel