搜索条件
1 - 2 共 2
每页显示结果数
搜索结果
-
- 描述:
- Identification and recognition of number plate is very difficult from low resolution images due to poor boundary and contrast. Our goal is to identify the digits from a low-quality number plate image correctly, but correct detection was exceedingly difficult in some cases due to the low-resolution image. Another goal of this paper was to upscale the image from a very low resolution to high resolution to recover helpful information to improve the accuracy of number plate detection and recognition. We have used Enhanced- Super-Resolution with Generative Adversarial Network (SRGAN). We modified native Dense Blocks of the Generative Adversarial Network with a Residual in Residual Dense Block model. In addition to Convolutional Neural Networks for thresholding. We also used a Rectified Linear Unit (ReLU) activation layer. The plate image is then used for segmentation using the OCR model for detection and recognizing the characters in the number plates. The Optical character recognition (OCR) model reaches an average accuracy of 84% for high resolution, whereas the accuracy is 4% - 7% for low resolution. The model’s accuracy increases with the resolution enhancement of the plate images. ESRGAN provides better enhancement of low-resolution images than SRGAN and Pro-SRGAN, which the OCR model validates. The accuracy significantly increased digit/alphabet detection in the number plate than the original low-resolution image when converted to a high-resolution image using ESRGAN.
- 关键词:
- Structural similarity of images, Number plate detection, Residual dense block, Super-resolution, Deep learning, and Optical character recognition
- 学科:
- Artificial Intelligence and Data Science
- 创造者:
- Roy, Sudipta, Ganguly, Debayan , Pal, Debojyoti , Chatterjee, Kingshuk , and Kabiraj, Anwesh
- 贡献者:
- Jio Institute, CVMIComputer Vision in Medical Imaging Project
- Owner:
- n.sakthivel@jioinstitute.edu.in
- 出版者:
- Springer Nature
- 位置:
- Switzerland
- 语言:
- English
- 日期上传:
- 11-02-2023
- 修改日期:
- 16-02-2023
- 创建日期:
- 01-09-2022
- Rights Statement Tesim:
- In Copyright
- License Tesim:
- All rights reserved
- Resource Type:
- Article
- 识别码:
- 10.1007/s11042-022-14018-0
-
- 描述:
- Fetal brain segmentation and gestational age prediction have been under active research in the field of medical image processing for a long time. However, both these tasks are challenging due to factors like difficulty in acquiring a proper fetal brain image owing to the fetal movement during the scan. With the recent advancements in deep learning, many models have been proposed for performing both the tasks, individually, with good accuracy. In this paper, we present Multi-Tasking Single Encoder U-Net, MTSE U-Net, a deep learning architecture for performing three tasks on fetal brain images. The first task is the segmentation of the fetal brain into its seven components: intracranial space and extra-axial cerebrospinal fluid spaces, gray matter, white matter, ventricles, cerebellum, deep gray matter, and brainstem, and spinal cord. The second task is the prediction of the type of the fetal brain (pathological or neurotypical). The third task is the prediction of the gestational age of the fetus from its brain. All of this will be performed by a single model. The fetal brain images can be obtained by segmenting it from the fetal magnetic resonance images using any of the previous works on fetal brain segmentation, thus showing our work as an extension of the already existing segmentation works. The Jaccard similarity and Dice score for the segmentation task by this model are 77 and 82%, respectively, accuracy for the type of prediction task is 89% and the mean absolute error for the gestational age task is 0.83 weeks. The salient region identification by the model is also tested and these results show that a single model can perform multiple, but related, tasks simultaneously with good accuracy, thus eliminating the need to use separate models for each task.
- 关键词:
- Medical image processing, Fetal brain segmentation, Deep learning, Fetal gestational age prediction, and Convolutional neural networks
- 学科:
- Data Science and Artificial Intelligence
- 创造者:
- Ganguly, Debayan , Chatterjee, Kingshuk , Gangopadhyay, Tuhinangshu , Sarkar, Surjadeep , Halder, Shinjini , Dasgupta, Paramik , and Roy, Sudipta
- Owner:
- n.sakthivel@jioinstitute.edu.in
- 出版者:
- Springer Nature
- 位置:
- Switzerland
- 语言:
- English
- 日期上传:
- 11-02-2023
- 修改日期:
- 16-02-2023
- 创建日期:
- 01-11-2022
- Rights Statement Tesim:
- In Copyright
- License Tesim:
- All rights reserved
- Resource Type:
- Article
- 识别码:
- 10.1007/s13721-022-00394-y