Test paper automatic processing system and method based on OpenCV

Abstract

The invention discloses a test paper automatic processing system and method based on OpenCV. The system is used for automatically processing a specified test paper, and comprises a test paper automatic identification unit based on OpenCV and a paper inspection management unit based on a B / S framework; the specified test paper is provided with a plural number of question number figures, a rectangular answer frame is arranged between each pair of adjacent question number figures, and a boundary area of the periphery of the specified test paper is provided with a plural number of question mark points used for marking positions of the question number figures; the test paper automatic identification unit comprises a test paper image acquisition and preprocessing module, a mark point processingand identification module, a question number identification and location module, an answer area identification and cutting module and a full-filling content identification module; the paper inspectionmanagement unit comprises a paper inspection logic control module, a paper inspection task distributing module, a user information management module, a content management module and a user interfacemodule. The test paper automatic processing system and method based on the OpenCV have the advantages that answers exceeding the answer area can be displayed, manual work is saved, and the paper inspection efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an OpenCV-based automatic test paper processing system and method. Background technique [0002] Traditionally, the examination papers are marked manually. When there are a large number of examination papers to be reviewed, the efficiency of the manual method not only becomes very low, but also prone to errors, which affects the evaluation results of the examination. Therefore, the online marking system for the examination papers in the form of answer sheets has emerged, which realizes automatic scoring through electronic scanning and image recognition technology, which greatly improves the efficiency of marking papers and reduces human misjudgments. [0003] At present, the widely used online marking system and the automatic marking technology of objective question answer sheets have become mature. However, there is still a lack of research on the direct identification...

AI Technical Summary

Problems solved by technology

However, there is still a lack of research on the direct identification of the complete test paper, especially the subjective part. At this stage, the administrator manually marks the subjective question answering area and creates the answer sheet template for marking.

There are many problems in this marking method. For example, the answering area cannot be dynamically adjusted, so that the marking teacher can only see the answers within the set area, and the answers beyond the area cannot be graded; the fixedly divided and saved answer pictures occupy a large number of servers Storage; Manually marking the answer area also adds a lot of workload to the test administrators, making it difficult to apply this technology to university exams with many test subjects

On the other hand, the existing online marking system generally adopts the model of test paper scanning server + marking management server + dedicated marking client.

Images