Research Of Intelligent Image Processing And Cloud Application Based On Salient Region Detection

The process from films into digital images has come to the end, and the time of bigdata is coming. Uses are creating uncountable amount of data every day, includingnumberous images and videos. The resolution of users’ equipment is becoming higher andhigher. They can produce about10G’s data in a trip. How to deal with such amount ofimages is a key problem that should be solved. Aiming at this user need, this thesis didsome research on the existing intelligent automatic processing technic and cloudcomputing realization method. Based on the salient region detection method, this thesisdesign a novel intelligent automatic processing method which has combined with cloudcomputing, significantly improved the existing processing method and helps users inimage shortage and processing.The thesis took the image auto-processing algorithm in Photoshop as the basementfor us to compare with and took the expose algorithm as the example. The SPK method isthe salient segmentation method that chose from the current salient detection methods bycomparing them. Combined with the salient segmentation and content analysis, the thesisimproved the existing auto-processing algorithm and made it intelligent. The thesis alsocompared the result between before and after improvement, and did analysis in multipleconditions. The thesis also analyzed the possibility of deploying this method on the Cloudside. It took the AWS as the Cloud base, the S3as the shortage base and the EC2as thecomputing base. The design of the user interface, the functional parts and the work flow ofthe method are also included in the thesis.Cloud computing can provide an efficient and fast computing which solved theproblem of occupying the computing resources and time, and offer the service in multipledevices which makes the method easy to use.