中文摘要：

为了提高用户-物体交互行为中可供性物体的识别效率和准确度，为产品形态的可用性设计提供有效的形态特征元素，提出了一种基于深度图像技术的可供性物体获取方法。该方法基于可供性构建了AHO（Action—Hand-Ohject）可供性模型和H0（Hand-Object）＋势，应用深度图像技术获取HO手势的表现特征向量，并训练支持向量机以识别HO手势，结合数理统计方法获取相应的可供性物体。对锅具设计中的两种HO手势识别实验中，该方法能有效地识别出相应可供性物体的形态特征元素。

英文摘要：

For improving the efficiency and accuracy of recognizing affordance objects during user-object interaction to provide effective shape feature elements for useful design of product shape, an qffordance object acquiring method based on depth image technology was proposed. This method constructed AH0 （Action-Hand-Object） affordance model and HO （Hand-Object） gesture based on affordance , applying depth image technology to acquire H0 gesture expression feature vector, training support vector machine to recognize H0 gesture, and combining statistic method to get corresponding affordance object. The recognizing experiments of two HO gestures for pan shows this method can efficiently recognize shape feature elements of corresponding affordance objects.