Abstrak
The authenticity of school assignments is crucial for maintaining the objectivity of learning evaluations. However, the practice of having assignments submitted by third parties is still common. This study develops an assignment authenticity verification system based on handwriting analysis using image processing and machine learning approaches. This system utilizes a Convolutional Neural Network (CNN) for visual feature extraction and a Siamese network to measure handwriting similarity. The data used includes the writing of sixth-grade students at SD Negeri 060868 Medan Timur and writing from non-students as comparisons. The handwriting images were converted to grayscale, resized to 128x128 pixels, and then processed to generate a one-dimensional feature vector. A similarity score was calculated using Euclidean distance, and if the value is ? 0.75, it is considered a match. This system was built with Python, TensorFlow, and OpenCV, and integrated into web and mobile platforms. Trial results show that the system is capable of automatically verifying assignment authenticity, helping teachers assess more objectively and efficiently, and improving academic transparency.
Referensi
Adak, C., Marinai, S., & Chaudhuri, B. B. (2018). Offline Bengali writer verification by PDF-CNN and Siamese net. Proceedings of the IEEE International Conference on Frontiers in Handwriting Recognition. https://doi.org/10.1109/DAS.2018.33
Afzali, P., & Rezapour, A. (2024). Leveraging deep feature learning for handwriting biometric authentication. Research in Intelligent and Embedded Systems, 13(1), 88-103. https://doi.org/10.22105/riej.2024.432510.1412
Ahrabian, K., & BabaAli, B. (2019). Usage of autoencoders and Siamese networks for online handwritten signature verification. Neural Computing and Applications, 31, 9321-9334. https://doi.org/10.1007/s00521-018-3844-z
Dlamini, N., & Van Zyl, T. L. (2019, November). Author identification from handwritten characters using Siamese CNN. In 2019 international multidisciplinary information technology and engineering conference (IMITEC) (pp. 1-6). IEEE. https://doi.org/10.1109/IMITEC45504.2019.9015897
Du, W., Fang, M., & Shen, M. (2017). Siamese convolutional neural networks for authorship verification. Stanford CS231n Report.
Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
Hadsell, R., Chopra, S., & LeCun, Y. (2006, June). Dimensionality reduction by learning an invariant mapping. In 2006 IEEE computer society conference on computer vision and pattern recognition (CVPR'06) (Vol. 2, pp. 1735-1742). IEEE. https://doi.org/10.1109/CVPR.2006.100
Jambhale, T., Sharma, L., Narayan, P., & Vijayarajan, V. (2021, December). One shot verification of handwritten signatures using Siamese networks. In 2021 International Conference on Advances in Computing, Communication, and Control (ICAC3) (pp. 1-7). IEEE. https://doi.org/10.1109/ICAC353642.2021.9697248
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, 25, 1097–1105.
LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521, 436–444. https://doi.org/10.1038/nature14539
Sharma, N., Gupta, S., Mohamed, H. G., & Anand, D. (2022). Siamese convolutional neural network-based twin structure model for independent offline signature verification. Sustainability, 14(18), 11484. https://doi.org/10.3390/su141811484
Suteddy, W., Agustini, D. A. R., & Atmanto, D. A. (2024). Offline handwriting writer identification using depth-wise separable convolution with Siamese network. JOIV: International Journal on Informatics Visualization, 8(1), 535-541. https://dx.doi.org/10.62527/joiv.8.1.2148
Vorugunti, C. S., Mukherjee, P., & Pulabaigari, V. (2019, September). OSVNet: Convolutional Siamese network for writer-independent online signature verification. In 2019 international conference on document analysis and recognition (ICDAR) (pp. 1470-1475). IEEE. https://doi.org/10.1109/ICDAR.2019.00236
Xiao, W., & Ding, Y. (2022). A two-stage Siamese network model for offline handwritten signature verification. Symmetry, 14(6), 1216. https://doi.org/10.3390/sym14061216
Zhang, Y., Wang, X., & Lin, D. (2019). Curriculum learning for Siamese neural networks in visual tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41(12), 2987–3001.
Artikel ini berlisensiCreative Commons Attribution-ShareAlike 4.0 International License.
Hak Cipta (c) 2025 Indah Rahayu, Fatma Sari Hutagulung