网络技术的发展和图像获取设备的普及导致数字图像迅速增长,依靠先进的技术提取图像蕴含的情感语义实现图像情感语义分类正是当前各行业急需解决的问题。为此提出一种基于改进的OCC情感模型的自然风景图像情感语义分类方法。通过融入性格、心情因素描述图像的个性情感,使用BP神经网络实现,解决图像分类中的语义理解问题。使用百度图片频道上下载的600张场景图像进行训练和测试,实验通过与人工计算结果相比较,取得了良好的分类效果,可为更多类型的图像情感语义分类打好基础,具有一定的实用价值。
The development of network technology and the popularisation of image acquisition devices result in the rapid growth of digital images. It becomes an urgent issue in various trades at present that to realise the emotion and semantics classification of images by extracting the emotional semantics implicated in the image relying on advanced technologies. Therefore, we propose an emotional semantics classification method for natural scene images which is based on improved OCC affective model. The method describes images' personality emotion by integrating the disposition and mood factors, uses BP neural network to implement it, and this solves the problem of semantic comprehension in image classification. We use 600 scene images downloaded from Baidu pictures channel to train and test the method, and the experiments achieve good classification effect in comparing it with the manual computing results. The proposed method can lay a good foundation for more types of emotional semantics classification of images and has certain practical value.