Design And Implementation Of Image Evaluation System Based On Android Platform

The concept of computational aesthetics was formally proposed at the first session of the visual computing aesthetics academic conference.Computational aesthetics has become a branch of the field of image processing.It is distinguished from the traditional image quality evaluation field of definition,resolution and noise.Computational aesthetics is more important to evaluate and optimize the image quality.Point from the bottom feature to the aesthetic angle of the high level.People try to propose an algorithm that can make computers recognize the beauty and ugliness of images.The research results of this algorithm will have a great prospect in the future,for example,map search,photography analysis,human-computer interaction and advertising design,so that the research in this field has become one of the hot topics in the field of computer vision.At present,the core functions of the mobile image processing APP focus on the beautification of the image and the beauty of the human image,and the beauty of the processed images is not beautiful,and the beauty to what extent,the APP does not give a unified standard and evaluation,which is entirely dependent on the user's understanding and definition of the beauty itself.This thesis applies the algorithm of image aesthetic quality assessment to engineering practice.The main research content of this thesis is to design an image processing APP based on aesthetic quality evaluation.Its core function is to give an aesthetic evaluation of an image using aesthetic quality evaluation algorithm selected in the mobile phone,and to adapt the image adaptively according to the evaluation standard,and make it become a higher quality and more aesthetic image.The beauty of image is not beautiful mainly depends on the subjective feelings of each dimension of the image.This thesis studies the aesthetic quality evaluation algorithm of each image,and tests the effectiveness and performance of each algorithm through experiments.After a large number of experiments and comprehensive analysis on the Photo.net Dataset image library,the quality evaluationof the image aesthetics is used in this thesis.The algorithm is effective and efficient.In this thesis,the algorithm is applied to the engineering of Android platform,and the engineering application value of the algorithm is realized.In addition to the core functions,the APP in this thesis has also designed functions such as image editing,puzzles,and image special effects,making the entire APP function more complete.It can basically satisfy the commercial standards.The implementation of the core algorithm of this APP isusing C++ combined with OpenCV.The running efficiency of the algorithm is also faster than we expected.