Image processing method and system based on AI mobile terminal automatic test framework

Abstract

The invention relates to an image processing method based on an AI mobile terminal automatic test framework. The image processing method comprises the steps of segmenting a page into multiple parts, extracting and processing sub-elements and features of each part, cutting the part with the similarity exceeding a threshold value, removing the page, generating layout diagrams, and fusing multiple layout diagrams; splitting a standard convolution of a Mobilenetv2 classification model into a deep convolution and a point-by-point convolution, and sequentially carrying out model structure calculation, memory efficient setting and Image Net classification; inputting a grey-scale image, an original image and a contour image of the image at the same time to carry out classification result prediction, and carrying out combination complementation on a plurality of output results; and inputting a screenshot of the picture and a data image layout file to obtain an lstm module, and compiling the lstm module into an automatic test script. Compiling cost and maintenance cost of the test case are greatly reduced, the cross-platform capability of the framework is improved, and the test case becomes more humanized.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image processing method and system based on an AI mobile terminal automated test framework. Background technique [0002] With the continuous efforts of the two major platforms of Android and IOS, the mobile operating systems on the market have been occupied by Android and IOS, and the share of Android is more than 80%. Then, in the face of various open source automated testing frameworks and tools on the market because of their own focus, there are more or less poor cross-platform capabilities, poor cross-application capabilities, greater dependence on ID for stability, high cost of control capture, dump Various problems such as the failure of the system view tree probability. [0003] The problems faced by major open source automated testing frameworks and tools are: such as UIAutomator: only supports android4.1 (API level 16) and above. Script recording i...

AI Technical Summary

Problems solved by technology

The library only supports using Java, so it is difficult to mix with Cucumber using Ruby

If you want to support the BDD framework, it is recommended to use Java's own BDD framework, such as Jbehave; Appium faces stability problems; Robotium cannot implement cross-platform and cross-application App testing; Espresso: Compared with Robotium and UIAutomator, it is characterized by a larger scale Smaller, more concise, more precise API, simple to write test code, easy to get started quickly, but because it is based on Instrumentation, it also cannot implement cross-App application testing

[0004] Therefore, the traditional framework has problems such as strong dependence on the system view tree, cumbersome locking of resource IDs and view types, and high maintenance costs for ID confusion.

Images