Research On Technology Of Automatically Annotating Images With Key Words

With the fast development and wide application of multimedia technology, computer technology, communication technology and Internet technology, people can collect and produce various kinds of multimedia data based on plenty of methods, and it is more and more important to organize and manage the multimedia information. Most frequent and maximal information of the multimedia data is the visual information. Therefore, the issue of the visual information retrieval has attracted many people's attention. For the non-structured image data, the traditional text-based image retrieval is so inefficient and thereby the content-based image retrieval is bought forward and it has developed a lot. However, there has a big hurdle in content-based image retrieval till the present moment, i.e.,"semantic gap"."Semantic gap"is the gulf between the low-level image visual feature and high-level concepts, the images can be different of semantic concepts while having similar visual features, and they can also be different of visual features while having the same concept. In this thesis, we proposed a solution to resolve the problem, which is auto image annotating technology based on machine learning method combined with the relations between the key words. The experiments''results proved the effectiveness of this method. The main researches in this paper are as following:Firstly, we choose the Normalized Cut algorithm to cut images into several areas in our experiments. We used the combination of color, texture and shape features as low-visual feature vector. The experiments'results proved that the combination of features can effectively improve the rate of recognition.Second, we contrast the features between the traditional BP algorithms and adopted the impulse factor one, and through the experiment's result, the impulse factor can effectively resolve local minimum value and improved the convergence.Third, because of the instability and the long training time of this network, we set the maximum interactive numbers of the unit in competitive layer to conquer the instability, and execute the learning algorithm in two phases to improve the effectiveness.Last, we proposed the method this combine the machine learning algorithm and the relations of image key words to automatically annotate the images. The experiments'results proved the combination of visual features and text relations can effectively improve the accuracy of image annotation.