论文部分内容阅读
为减小人为因素对三角方向鉴别测量的影响,有效提高测量效率,研究了以机器视觉系统替代观察者进行三角方向鉴别测量的途径,建立了一种专门针对三角方向鉴别测量的客观判别模型。该模型以人眼视觉系统对信号的处理过程以及人眼视觉系统的主要效应为基础,利用方向通道滤波器实现图像信号分解,通过计算视觉特征矢量相关度,实现三角形靶标方向的判别。结合凝视型成像系统虚拟样机提供的三角形靶标,进行了客观判别模型与观察者的判别性能对比实验,并给出三组不同靶标尺寸下的对比实验结果。实验结果表明,客观判别模型很好地预测了观察者的判别性能。
In order to reduce the influence of human factors on the triangulation direction measurement and improve the measurement efficiency effectively, the way of using the machine vision system to substitute the observer for the triangulation direction measurement was studied, and an objective discriminant model specifically aimed at the triangulation direction discrimination measurement was established. Based on the signal processing and the main effects of the human visual system, the model uses the directional channel filter to decompose the image signal. By calculating the correlation degree of the visual feature vector, the model determines the direction of the triangle target. Combined with the triangular target provided by the virtual prototyping system of gaze-type imaging system, the discriminative performance comparison between the objective discriminant model and the observer was carried out. The comparative experimental results of three different target sizes were given. Experimental results show that the objective discriminant model predicts the discriminant performance of observers.