ZHU Yonghong, XIONG Meng, ZHAO Yifeng, WANG Wei
( School of Mechanical & Electronic Engineering, Jingdezhen Ceramic Institute, Jingdezhen 333403, Jiangxi, China )

Abstract:In the firing process of ceramic products, the sintering conditions vary from phase to phase. In different firing phases, flame texture changes obviously, so it can be used as an important parameter for identifying the operating conditions of the firing zone in a ceramic kiln.K-means clustering-based color segmentation method of flame image in the firing zone for ceramic kiln is proposed in this paper, and at the same time the corresponding k-means clustering algorithm is also given. The effective segmentation of flame image sample for ceramic kiln is realized by color segmentation method with K-means clustering algorithm, and color information is reserved to an extreme. The experimental results show that this method will provide feature extraction of firing zone flame image for ceramic kiln with a good technical means. This will provide a new way for utilizing the flame image features to detect the firing zone temperature of ceramic kiln, so as to realize intelligent temperature control of ceramic kiln.
Key words:ceramic kiln; K-means clustering; flame image segmentation; color space