License Plate Image Super-Resolution Based on Convolutional Neural Network

Yang, Yang; Bi, Peng; Liu, Ying

doi:10.1109/icivc.2018.8492768

Cited by 18 publications

(10 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The researcher Yang [12], proposed a super-resolution method using a multi-scale super-solution convolutional neural network, with the motivation of the architecture Inception do GoogleNet. In addition, different sizes of convolution filters are used in order to achieve different characteristics of the image with low resolution.…”

Section: Related Workmentioning

confidence: 99%

“…In addition, different sizes of convolution filters are used in order to achieve different characteristics of the image with low resolution. The results obtained by Yang [12] considered several experiments, however, we will emphasize Yang's multi-scale super-resolution convolutional neural network (MSRCNN). The experiments cover the variation of the CNN layers, for different filter sizes, thus, the variations were from MSRCNN1 to MSRCNN8, besides that, the author presented all the results, however, in a table, highlighting all the variations of the MSRCNN method.…”

Section: Related Workmentioning

confidence: 99%

“…So, after the SR block, the video resolution was changed to 1920×1080 pixels. Thus, in this research, we use the bilateral total variation super-resolution algorithm (BTV-L1) [12].…”

Section: B Bilateral Full-range Super-resolutionmentioning

confidence: 99%

See 2 more Smart Citations

License Plate Detection System in Uncontrolled Environments using Super Resolution and a Simple Cell Phone Camera attached to a Bicycle Seat

Silva¹,

LGJ²,

Colares³

et al. 2022

Anais Do XL Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

In this paper, a low resolution Brazilian license plate detection system based on the TV-L1 bilateral superresolution method is proposed. The system is used in uncontrolled environments whose acquisition source is performed in motion to detect license plates using a point of view of a bicycle and a simple cell phone camera. The system is divided in: (i) super-resolution, in which the use of bilateral images from super-resolution videos TV-L1 was used to increase the spatial resolution of the designed database and; (ii) license plate detection, in which digital image processing techniques were used to convert the input image to grayscale and isolate the object of interest, through the morphological operation of closing and the algorithm of flood fill. As a result, the TV-L1 bilateral super-resolution method had an accuracy of 93%, making it a promising approach for future research.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

License Plate Detection System in Uncontrolled Environments using Super Resolution and a Simple Cell Phone Camera attached to a Bicycle Seat

Silva¹,

LGJ²,

Colares³

et al. 2022

Anais Do XL Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

show abstract

“…The Perceptual loss function consists of two losses adversarial and content loss (2) The above equation shows that the perceptual loss is the sum of the both losses. 1) Content Loss: The content loss is described by the Equation 3.…”

Section: B Perceptual Loss Functionmentioning

confidence: 99%

Super Resolution of Videos using SRGAN

Shah¹

2020

IJRASET

View full text Add to dashboard Cite

The primary objective of our project is to generate the Super resolved images. To enhance the quality of images we are using SRGAN techniques. There are various methods of image transformation where the computing system receives some input and transforms it in output image. Previous techniques of up scaling were based on minimizing the mean squares reconstructed error. Generative Adversarial Networks are the deep neural network architectures comprised of two networks Generator and Discriminator. GANs are about creating, like drawing a portrait or composing a symphony. SRGAN provides several benefits over other techniques it proposes a perceptual loss function which consists of adversarial loss and content loss. The two main blocks are Generator and Discriminator. The Discriminator discriminates between real HR images from generated super resolved images. The Generator function is to train a propagating model. Adversarial loss function uses discriminator network which is already trained to discriminate between the two images. The content loss function uses perceptual similarity instead of pixel space similarity. One of the best things about SRGANs is that they generate data that is similar to real data. SRGANs learn the internal representations of data to generate the upscale images. The neural network is successful in recovering the photo realistic textures from downgraded images. The SRGAN techniques lack high peak to signal noise ratio but provides high efficiency and visual perception. Combining the perceptual and adversarial loss will generate a high quality super resolved image. During training phase perceptual losses measure image similarities more robustly than per-pixel losses. Perceptual loss functions measures the high level perceptual and semantic differences between the images. Our method uses semantic content losses which has a VGG network and an edge promoting adversarial losses for edges. I.

show abstract

“…This classification is performed by Lung CT image recognition. In addition, ZFNet [11], VGGNet [12] and GoogleNet [13] are proposed. Deep Learning has been applied to remote semantic image segmentation.…”

Section: Introductionmentioning

confidence: 99%

Segmentation of Small Objects in Satellite Imagery Using Dense U-Net in Massachusetts Buildings Dataset

Haq

Arymurthy

Fadhil

2022

J. RESTI (Rekayasa Sist. Teknol. Inf.)

View full text Add to dashboard Cite

Class imbalance is a serious problem that disrupts the process of semantic segmentation of satellite imagery in urban areas in Earth remote sensing. Due to the large objects dominating the segmentation process, small object are consequently limited, so solutions based on optimizing overall accuracy are often unsatisfactory. Due to the class imbalance of semantic segmentation in Earth remote sensing images in urban areas, we developed the concept of Down-Sampling Block (DownBlock) to obtain contextual information and Up-Sampling Block (UpBlock) to restore the original resolution. We proposed an end-to-end deep convolutional neural network (DenseU-Net) architecture for pixel-wise urban remote sensing image segmentation. this method to segmentation the small object in satellite imagery.The accuracy of the small object class in this study was further improved using our proposed method. This study used data from the Massachusetts Buildings dataset using Dense U-Net method and obtained an overall accuracy of 84.34%.

show abstract

License Plate Image Super-Resolution Based on Convolutional Neural Network

Cited by 18 publications

References 8 publications

License Plate Detection System in Uncontrolled Environments using Super Resolution and a Simple Cell Phone Camera attached to a Bicycle Seat

License Plate Detection System in Uncontrolled Environments using Super Resolution and a Simple Cell Phone Camera attached to a Bicycle Seat

Super Resolution of Videos using SRGAN

Segmentation of Small Objects in Satellite Imagery Using Dense U-Net in Massachusetts Buildings Dataset

Contact Info

Product

Resources

About