Research On Techniques Of Image Retrieval At Education Resource Construction

Electronic information technology has been rapid developed since the beginning of the 21 th century. Along with the popularity of the Internet, multimedia resources have been greatly enriched, which brings great convenience to people's work and study. Meanwhile, people's learning materials tend to be more and more electronic as mobile learning and ubiquitous learning develop.Today, network education resources have already become an important part of learning resources. However, there are many kinds of network resources and whose type is not consistent. It is worth to discuss that how to effectively integrate these teaching resources together and let users have a good experience so that they can manage and use the platform in a convenient way. At the same time, images as a important information carrier takes a more and more big proportions in education resources. But with the explosive growth of network resources, how to find the target resources accurately to improve the efficiency and service quality has become an urgent problem to be solved in the construction of teaching resource platform.Focusing on the above problems, this paper has mainly done the following three aspects works:First, we discussed the necessity of the construction of teaching resources platform. After reading many relevant literature, we concluded its development in recent years and analyzed the problems that often occur in this process, then introduced the teaching theory it based on. Finally, based on full analysis of user's needs, a design scheme of teaching resources platform is proposed.Second, We discussed the significance of image retrieval technology in the construction of teaching resources platform. Then, the key technologies of content-based retrieval are introduced. Next, this paper studied an image retrieval method that based on comprehensive color and edge features. Aiming at the drawbacks of manual feature selection, we proposed the idea of applying Convolutional Neural Network to image retrieval techniques. Compared with thetraditional method of extracting the underlying feature of the image, the method that based on the convolution neural network can put image into classification model directly, and the neural network can automatically learn the feature of the image,reducing the interference of human factors on the classification results. At the same time, the unique network structure of Convolutional Neural Network can reduce the computational complexity of the algorithm and tolerate some degree of image distortion.Third,We combined Matlab with C#.NET,and realized the application of image retrieval technology based on the convolution neural net work in education resources platform.