期刊名称：Lecture Notes in Engineering and Computer Science
印刷版ISSN：2078-0958
电子版ISSN：2078-0966
出版年度：2018
卷号：2233&2234
页码：330-333
出版社：Newswood and International Association of Engineers
摘要：In these days, the integration of technology in
teaching-learning process has become a central role in order to
redesign a quality education system especially for the
development of interactive education. In this concern,
technology based analysis on the interaction between students
and teacher and the feedback of the students play key roles.
Thus, in this paper, we proposed the automatic students’
feedbacks analyzing system for the purpose of speeding up the
communication between students and teacher in the classroom
by using the image processing and deep learning technology. In
the proposed system, the students can use the five kind of color
cards for answering the questions or for describing their
feedbacks. Then the automatic students’ feedback analyzing
system will analyze the color cards objects by using the camera
and describe the analyzed result to the teacher. In this way, the
interaction between students and teacher can be faster and can
give a lot of benefit for the education system. In order to
implement this system, firstly, the color objects segmentation is
performed over the input image using the predefined color
thresholds. Then, the noise objects are removed by using the
predefined maximum size and minimum size thresholds. Finally,
the Deep Convolutional Neural Network (DCNN) is applied in
order to classify the five color cards objects and non-card color
objects. The experiments are performed on the image that have
been taken in the large classroom under the different
illumination condition. According to the experimental results,
the proposed system can robustly analyze the color cards objects
with the accuracy of 97.02% on the training data and 94.38%
for the testing data. The proposed system can give the
ubiquitous (anytime, anywhere) analyzing of the students’
feedback in the classroom.
关键词：students’ feedback; e-learning; automatic;
analyzing; deep learning
