计算机视觉检测是指利用计算机图像处理系统来模拟人类视觉进行检测对象的分类、识别和测量的技术.使用基于模糊聚类的图像RGB分量线性变换后的正交彩色特征,建立了应用于火焰检测的一种混合检测模型,并将其应用于视频序列图像中火焰的自动检测.实验结果验证了该方法的有效性.
Computer vision detection is a technology of simulating human vision to classify, identify and detect objects by using computer image processing systems. In this paper, combined with fuzzy clustering analysis, a hybrid color model is established to detect flame. The model is based on the features that can be achieved from the linear transformation of RGB of the image color. Experimental results conducted on the fire images demon- strate the effectiveness of the proposed method.