Automatic Recognition Method for Checkbox in Data Form Image
Zhang Shengnan, Yuan Shanlei, Niu Lianqiang · 2014
The checkbox recognition is the key question in the automatic processing of data form based on image recognition. This paper proposes an automatic recognition method for checkbox, which can effectively solve the problems, such as border rupture, adhesion of table lines with handwritten symbol. First of all, this method uses the maximum tilt angle of table-run for fast skew detection and correction of form image, then adopts the connected domain labeling algorithm to rapidly locate the checkbox and extract the standard image of handwritten symbol. Finally the structuhandwritten symbol recognitionre features of the above standard image are integrated to realize handwritten symbol recognition. The experiments show that the proposed method has high quality and efficiency of recognition.