Kawi Character Recognition on Copper Inscription Using YOLO Object Detection

Santoso, Rachmat; Suprapto, Yoyon K.; Yuniarno, Eko Mulyanto

doi:10.1109/cenim51130.2020.9297873

Cited by 6 publications

(9 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

Section: Literature Reviewmentioning

confidence: 99%

“…According to previous research, you only look once (YOLO) object detection methods are used in [5], [6], which implements the CNN architecture. YOLOv3-tiny was able to recognise the Kawi character on copper inscriptions due to its high detection accuracy (average of 97.93% in [5] and high detection speed. Meanwhile, the Oracle Bone inscriptions (OBIs) were recognised using two deep learning models in [6]: first, YOLOv3-tiny was used to detect and recognise OBIs, and second, MobileNet was used to detect undetected OBIs, as YOLOv3-tiny's limitations prevent all OBIs from being correctly recognised.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Recognition and Translation of Ancient South Arabian Musnad Inscriptions

Altalhi,

Alwethinani,

Alghamdi

et al. 2023

IJACSA

View full text Add to dashboard Cite

Inscriptions play an important role in preserving historical information. As such, conservation of these inscriptions provides valuable insights into the history and cultural heritage of the region. Musnad inscriptions are considered one of the earliest forms of writing from the Arabian Peninsula, preceding the modern Arabic font; however, most Musnad inscriptions remain unread and untranslated, signifying a substantial loss of historical information. In response, this paper represents a significant contribution to the field by proposing a successful approach to interpreting Musnad inscriptions. To do so, a dataset was prepared from the Saudi Arabian Ministry of Culture and subjected to preprocessing for optimal recognition, a step that entailed several experiments to enhance image quality and preparedness for recognition. The dataset was then trained and tested with 29 classes using three different convolutional neural network (CNN) architectures: Visual Geometry Group 16 (VGG16), Residual Network 50 (ResNet50) and MobileNetV2. Thereafter, the performance of each architecture was evaluated based on its accuracy in recognising Musnad inscriptions. The results demonstrate that VGG16 achieved the highest accuracy of 93.81%, followed by ResNet50 at 89.39% and MobileNetV2 at 80.02%.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Section: Literature Reviewmentioning

confidence: 99%

Recognition and Translation of Ancient South Arabian Musnad Inscriptions

Altalhi,

Alwethinani,

Alghamdi

et al. 2023

IJACSA

View full text Add to dashboard Cite

show abstract

“…Yang et al [10] proposed a recognition guided detector (RGD) method that achieves tight Chinese character detection in historical documents. Santoso et al [11] detected Kawi characters using the YOLO architecture. The experimental results of this study demonstrated that the proposed method achieved the highest accuracy compared to other methods.…”

Section: Related Workmentioning

confidence: 99%

LONTAR_DETC: Dense and High Variance Balinese Character Detection Method in Lontar Manuscripts

2022

View full text Add to dashboard Cite

This paper proposed LONTAR_DETC, a method to detect handwritten Balinese characters in Lontar manuscripts. LONTAR_DETC is a deep learning architecture based on YOLO. The detection of Balinese characters in Lontar manuscripts is challenging due to the characteristics of Balinese characters in Lontar manuscripts. Balinese characters in Lontar manuscripts are dense, overlapping, have high variance, contain noise, and classes of these characters are imbalanced. The proposed method consists of three steps, namely data generation, Lontar manuscript annotation, and Balinese character detection. The first step is data generation, in which synthetic images of original Lontar manuscript images are generated with enhanced image quality. The second step is data annotation to build a new Lontar manuscript dataset. As a result, we also propose the Handwritten Balinese Character of Lontar manuscript (HBCL_DETC) dataset, a novel Balinese character in Lontar manuscripts dataset. HBCL_DETC contains 600 images that consists of more than 100,000 Balinese characters annotated by experts. Finally, the third step is training the YOLOv4 detection model using the HBCL_DETC dataset. We created this dataset specifically for the task of detecting Balinese characters in Lontar manuscripts. To evaluate the reliability of the dataset, we experimented with three scenarios. In the first scenario, the detection model was trained using original images of Lontar manuscripts, in the second scenario the detection model was trained with the addition of augmented grayscale images, and in the third scenario the detection model was trained using HBCL_DETC. Based on the experimental results, LONTAR_DETC can detect Balinese characters with high detection rate with mAP of 99.55%.

show abstract

“…This was the case in Bui, et al, 2016, where their CNN was trained on RGB images but tested on RGB converted greyscale images and found that the greyscale test images yielded an improvement of +1.4% in detector accuracy over the same RGB test images (Bui, et al, 2016). In addition, because greyscale lacks colour information, it requires less computing memory to process convolutional calculations (Bui, et al, 2016;Santoso, Suprapto, & Yuniarno, 2020) and allows for faster training time (Ng, Tay, & Goi, 2013).…”

Section: Neural Network Training With Greyscale Imagesmentioning

confidence: 99%

“…There have also been studies using greyscale images to train CNNs. In Santoso, et al, 2020, YOLOv3-Tiny was trained on greyscale images and used RGB as test images to detect copper inscriptions which yielded a high average detection accuracy of 97.93% (Santoso, Suprapto, & Yuniarno, 2020). An investigation completed by Ng, et al, 2013, compared training CNNs with greyscale, RGB and YUV images, and it was found that training on greyscale images produced the lowest detector error rate (Ng, Tay, & Goi, 2013).…”

Section: Neural Network Training With Greyscale Imagesmentioning

confidence: 99%

Near Infrared Imaging and Image Pre-Processing to Improve the Automatic Detection of Canada Geese

Szeto¹

View full text Add to dashboard Cite

Migratory shorebirds populations are adversely affected by climate change and loss of habitat thus careful monitoring of their populations is important for early detection of population loss. Current counting methods generally rely on intrusive and time-consuming manual identification. This work is part of a larger project to develop automated classification and counting methods using a remotely piloted aircraft system (RPAS). In addition to the use of RPAS, this work will also investigate if near-infrared (NIR) imaging captured by the RPAS yields detection improvements. Healthy vegetation reflects NIR wavelengths of light which can potentially create a greater contrast between an object and the surrounding vegetation. Preprocessing NIR raw images to enhance the contrast between vegetation and Canada geese (Branta canadensis) to improve object detection using the convolutional neural network (CNN) YOLOv4-Tiny have been investigated in this study. Training was done on a small dataset (423 and 1,269 images), a large dataset (2,000 images), artificially generated dataset (2,000 images) and hybrid datasets consisting of real and artificially generated images of Canada geese (4,000 and 5,000 images). A RPAS was used to obtain greyscale and NIR test images of geese decoys using the RPAS onboard camera and a NIR specific camera at varying altitudes. The NIR preprocessed ground test images showed detection improvements in both number of detections and confidence score percentages when validated against the YOLOv4-Tiny detector that was trained on an augmented dataset and the hybrid datasets. However, the greyscale aerial test images generally outperformed the NIR pre-processed images.

show abstract

Kawi Character Recognition on Copper Inscription Using YOLO Object Detection

Cited by 6 publications

References 10 publications

Recognition and Translation of Ancient South Arabian Musnad Inscriptions

Recognition and Translation of Ancient South Arabian Musnad Inscriptions

LONTAR_DETC: Dense and High Variance Balinese Character Detection Method in Lontar Manuscripts

Near Infrared Imaging and Image Pre-Processing to Improve the Automatic Detection of Canada Geese

Contact Info

Product

Resources

About