Classification of Image Distortions Based on Features Evaluation

Alaql, Omar; Ghazinour, Kambiz; Chang, Cheng

doi:10.1109/ism.2016.0076

Cited by 5 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhai and Min [14] provided an overview of classical algorithms and recent progresses in perceptual image quality assessment. Alaql et al [15] investigated the performance of different classification techniques and features to improve classification of distortions followed by IQA. Based on neural network, Kaur A et al proposed a novel no-reference IQA method using canny magnitude and achieved excellent results on LIVE and TID2008 datasets, proving the efficiency of ANN [13].…”

Section: Related Work and Novelties And Necessity Of The Studymentioning

confidence: 99%

A Unified Framework for Anomaly Detection of Satellite Images Based on Well-Designed Features and an Artificial Neural Network

Wang

You

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

Nowadays, as the number of remote sensing satellites launched and applied in China has been mounting, relevant institutions’ workload of processing raw satellite images to be distributed to users is also growing. However, due to factors such as extreme atmospheric conditions, diversification of on-board device status, data loss during transmission and algorithm issues of ground systems, defect of image quality is inevitable, including abnormal color, color cast, data missing, obvious stitching line between Charge-Coupled Devices (CCDs), and inconstant radiation values between CCDs. Product application has also been impeded. This study presents a unified framework based on well-designed features an Artificial Neural Network (ANN) to automatically identify defective images. Samples were collected to form the dataset for training and validation, systematic experiments designed to verify the effectiveness of the features, and the optimal network architecture of ANN determined. Moreover, an effective method was proposed to explain the inference of ANN based on local gradient approximation. The recall of our final model reached 81.18% and F1 score 80.13%, verifying the effectiveness of our method.

show abstract

Section: Related Work and Novelties And Necessity Of The Studymentioning

confidence: 99%

A Unified Framework for Anomaly Detection of Satellite Images Based on Well-Designed Features and an Artificial Neural Network

Wang

You

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…In [23, 24], we proposed an image distortion classification model that presents an efficient set of features which overcome the limitations of existing blind IQA features with regard to representing different distortion types and mixtures. In that framework [23, 24], a total of 30 features, which have been validated to provide significant information about different distortion types of an image, were selected as input to the deep learning model. These features are more efficient than many other state‐of‐the‐art blind IQA features.…”

Section: Our Approachmentioning

confidence: 99%

“…The RBM consists of m visible units

v = ()(, v_{1}, v_{2}, \dots, v_{m})

and n hidden units

h = ()(, h_{1}, h_{2}, \dots, h_{n})

, but there are no visible‐to‐visible‐ and hidden‐to‐hidden‐connections [25]. The total of 30 statistical features that have been validated in [23, 24] to contribute significant information regarding distortion visibility were selected as input to each DBN. The DBN learns the relationship between image representation and labels in two phases; unsupervised pre‐training and supervised fine tuning.…”

Section: Our Approachmentioning

confidence: 99%

See 1 more Smart Citation

No‐reference image quality metric based on multiple deep belief networks

Alaql

2019

IET Image Processing

Self Cite

View full text Add to dashboard Cite

“…These probabilities are used as weight values to indicate the amount of each distortion in the image. In this work, we use our classification model that is proposed in [2] [3]. In the regression stage, the final quality score based on five trained models; one model for each distortion type.…”

Section: The Framework Structurementioning

confidence: 99%

No-reference image quality metric based on features evaluation

Alaql

Ghazinour

2017

2017 IEEE 7th Annual Computing and Communication Workshop and Conference (CCWC)

Self Cite

View full text Add to dashboard Cite

Abstract-We develop an efficient general-purpose noreference (NR) image quality assessment (IQA) model that based on a hypothesis that an effective combination of image features can be used to develop NR-IQA approaches having competitive performance with the state-of-the-art. First, we design a collection of features, then evaluate the usefulness of each feature on different kinds of distortions using different features evaluation techniques. Therefore, we came up with a set of optimal features for each learning model in the framework. Our experimental results show that our approach outperforms state-of-the-art blind image quality prediction models.

show abstract

Classification of Image Distortions Based on Features Evaluation

Cited by 5 publications

References 17 publications

A Unified Framework for Anomaly Detection of Satellite Images Based on Well-Designed Features and an Artificial Neural Network

A Unified Framework for Anomaly Detection of Satellite Images Based on Well-Designed Features and an Artificial Neural Network

No‐reference image quality metric based on multiple deep belief networks

No-reference image quality metric based on features evaluation

Contact Info

Product

Resources

About