Research On Robot Calligraphy Based On Image Recognition

As a kind of important information recording method, the writing brush calligraphy played a huge role in the inheritance and development of the Chinese traditional culture for thousands of years. Early twenty-first Century, calligraphy was included in "the representative of the intangible cultural heritage of mankind" by the United Nations. It shows that the calligraphy is an important part of Chinese traditional culture. In the network era, more and more attention is not attached to the calligraphy, it needs to be inherited and carried forward, so popularization of calligraphy and calligraphy is the key to promote the current. However, the serious shortage of teachers in the current writing brush calligraphy has become one of the bottlenecks in the popularization of calligraphy and calligraphy. Industrial robots can imitate people do all kinds of work, and even the fine hand movements, has the potential to be used for calligraphy and calligraphy teaching. This paper focuses on the research of robot calligraphy with image processing methods. It is expected to provide an important way to ease the lack of calligraphy teacher problem. The work and results of this paper are as follows:(1) The parametric description of the calligraphy is studied, and the image recognition algorithm is designed. Among them, the writing brush calligraphy is divided into two levels:stroke and character. According to the characteristics of the description parameters, design the image recognition algorithm. "Point" uses "water drop" model to describe. Stroke is divided into simple strokes and strokes, including strokes, width, length and so on. The text is mainly described by the number of strokes, intersection location and intersection angle and other parameters.(2) The relationship between the control parameters of the machine writing and the appearance of the stroke and the stroke is established through experiments. A point of writing model is established by the regression model between the point of the "water drop" parameter and the pressing depth of the writing brush. Strokes of the writing are established on the respect of the width of the stroke, the robot's writing speed, and the depth of the regression model.(3) Robot calligraphy system is built based on IRB120 ABB robot and robot calligraphy software is developed. The software can identify the calligraphy description parameters and control the robot writing from the calligraphy image. Robot calligraphy and creative experiment was completed, verify the effectiveness of the proposed method.