Food Packaging OCR Dataset

Food packaging OCR dataset aims to study whether the field-specific data set can improve the text detection and recognition performance of the OCR model on food packaging images. The dataset contains annotation data for text detection and text recognition, covering product name, brand, ingredient list, nutritional information and validity period. The image is collected from the real food packaging and contains different lighting, shooting angles and packaging design conditions, so it can reflect the actual application scenario. This dataset can be used to train, evaluate and compare OCR models.