Retrieval of Design Patent Images Based on MapReduce Model
Jiangzhong Cao · Journal of Chinese Computer Systems · 2012
In the field of industrial design,the demand for intellectual property protection has been increasing.To settle the conflict between the rapid growth of the data which is in the database of patent design image retrieval and the demand for real-time retrieval,we combine the content-based image retrieval technology with MapReduce framework,develop the intellectual content-based image retrieval system.The mission of Map is to extract the feature of shape,texture and color in the retrieval image,which is used to match with the features in the database of patent feature.The mission of Reduce is to sort the matching result according to the similarity,and finally get the retrieval result.The result of our experiment shows that the algorithm we proposed can not only balance the load of system,but also can increase the utilization ratio of the resources,which significantly decrease the retrieval time in the large data sets.