Method and device for classifying urban management cases

Abstract

The invention relates to a method and a device for classifying urban management cases. The method comprises steps: urban management case samples are acquired, and sample images corresponding to the urban management case samples are extracted; brightness and contrast normalization processing is carried out on the sample images; ZCA whitening processing is carried out on the sample images through brightness and contrast normalization processing; according to the sample images after ZCA whitening processing, a convolutional neural network CNN model is built; and according to the CNN model, images corresponding to the inputted to-be-classified urban management cases are classified to divide the to-be-classified urban management cases according to the classification result. According to the method and the device for classifying urban management cases, case images in an intelligent urban management application system are classified, and the method and the device have the advantages of short classification time, high classification precision, low application cost, simple model structure and the like.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and device for classifying urban management cases. Background technique [0002] The smart urban management system is an integrated system for reporting and handling urban management cases developed using "Internet +" technology, which includes PC terminals and mobile terminals. As the number of users of the smart urban management application system increases, more and more cases are reported to the system. According to statistics, the number of reported cases per day is currently around several thousand. Each case has at least one case picture, which causes a problem that a large number of cases cannot be automatically classified, that is, given a case picture, it is impossible to determine which type of case it belongs to. [0003] Although some images can be classified using the traditional Convolutional Neural Network (CNN) algorithm or Support Vector Mach...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the defects that traditional CNN algorithm or SVM algorithm etc. are used for image classification and cannot meet the requirement of case classification accuracy, the present invention proposes the following technical solutions:

Images