Our research group has been investigating various aspects of graphics recognition techniques for more than ten years. We have worked on map analysis, symbol recognition, dimension analysis and the conversion of engineering drawings to CAD models. Lately, we are also conducting research on the interpretation of architectural drawings. In addition, we have built up a software platform of generic tool for graphics document image analysis, and we have participated in many international activities around the topic of graphics recognition. In this paper, we present some of our achievements and results from these ten years, and we propose a number of open problems, which we think are good challenges in the coming years, for ourselves and for other teams.