模糊图像分类

Fuzzy Image Classification

图像质量对视觉系统来说至关重要，但由于技术水平的有限性，导致图像在获取、传输以及存储过程中会模糊失真。模糊图像分类是模糊图像复原获取高质量图像的前提条件，是视觉系统的重要组成，所以，模糊图像分类具有重要的研究意义。

Image quality is very important to the visual system. However, due to the limited level of technology, the image will be blurred and distorted during acquisition, transmission and storage. Fuzzy image classification is the prerequisite for the restoration of high quality images as well as the important component of the high quality visual system. Therefore, fuzzy imageclassification has wide application prospects.

本实验室分别利用传统的机器学习方法和深度学习方法实现模糊图像的模糊类型快速准确的识别和分类。在此过程中本实验室建造了自己的模糊图像数据库，分别包含了仿真模糊图像库和自然模糊图像库。

Using statistical learning method and deep learning method, we have realized the identification and classification of the fuzzy types for fuzzy images [paper]. In this process, our laboratory has built our own blur image database, named as BHBID (Beihang University Blur Image Database), which includes simulation blur image database and natural blur image database.

数据库中所有数据如附件中压缩文件，数据库中的样本图像如下：

All the data in the database is compressed in the attachment. The sample images in the database are as follows: