OCR recognition model training method and device based on crowdsourcing technology and computer equipment

The invention discloses an OCR recognition model training method based on a crowdsourcing technology. The OCR recognition model training method comprises the steps: recognizing a to-be-recognized image according to a basic OCR recognition model, and transmitting a recognition result to a crowdsourcing system; receiving a verification result fed back by each crowdsourcing terminal of the crowdsourcing system for the identification result; judging whether the recognition result is passed or not, and recording the to-be-recognized image and the verification result when the recognition result is not passed; and finally, retraining the basic OCR model according to a verification result to obtain a target OCR model. The invention further provides an OCR recognition model training device based onthe crowdsourcing technology, computer equipment and a computer readable storage medium. According to the OCR recognition model training method, the basic OCR model can be retrained through the crowdsourcing verification result under the condition that a large amount of sampling data is not needed, so that the accuracy of the OCR model is improved, and the training efficiency is high, and the cost is low.