Bio: Ankita Singh is an academic researcher. The author has contributed to research in topics: Histogram matching & Histogram equalization. The author has an hindex of 1, co-authored 1 publications receiving 3 citations.
••10 Mar 2016
TL;DR: A new algorithm for contrast improvement is proposed that reconstructs the enhanced image by applying the inverse DWT and results indicated that the image contrast enhanced by the purposed method was higher than that of the imagesEnhanced by the other conventional state-of-the-art techniques.
Abstract: Digital Imagery systems are traditionally bad in low light conditions. In this paper, a new algorithm for contrast improvement is proposed. The algorithm consists of two stages. The first stage is decomposing the input image into four subbands by applying two-dimensional discrete wavelet transform and estimates the singular value matrix of sub band image. The second stage is that it reconstructs the enhanced image by applying the inverse DWT. The technique is compared with conventional image equalization technique such as standard General Histogram Equalization (GHE) and other state-of-the-art techniques such as Quadrant Dynamic Histogram Equalization (QDHE), Singular-Value-Wavelet based image Equalization (SVWE) and Singular Value Equalization (SVE) on the basis of their Peak Signal to Noise Ratio (PSNR) and Root Mean Square Error (RMSE) values. The simulation results indicated that the image contrast enhanced by the purposed method was higher than that of the images enhanced by the other conventional state-of-the-art techniques.
TL;DR: A new classification of the main techniques of low-light image enhancement developed over the past decades is presented, dividing them into seven categories: gray transformation methods, histogram equalization methods, Retinex methods, frequency-domain methods, image fusion methods, defogging model methods and machine learning methods.
Abstract: Images captured under poor illumination conditions often exhibit characteristics such as low brightness, low contrast, a narrow gray range, and color distortion, as well as considerable noise, which seriously affect the subjective visual effect on human eyes and greatly limit the performance of various machine vision systems. The role of low-light image enhancement is to improve the visual effect of such images for the benefit of subsequent processing. This paper reviews the main techniques of low-light image enhancement developed over the past decades. First, we present a new classification of these algorithms, dividing them into seven categories: gray transformation methods, histogram equalization methods, Retinex methods, frequency-domain methods, image fusion methods, defogging model methods and machine learning methods. Then, all the categories of methods, including subcategories, are introduced in accordance with their principles and characteristics. In addition, various quality evaluation methods for enhanced images are detailed, and comparisons of different algorithms are discussed. Finally, the current research progress is summarized, and future research directions are suggested.
f< span>d Span>ise.
Abstract: Riceisastaple f o o din m ost o ft h e A sianco u n tries.Itisani m p o rta n tcro p , and o v e r h alf o ft h e w orld p o p u l ationrelies o nit f or f o o d.H o w eve r, p ad d y le a f d ise a se c anaffe c t bo th t he q u alitya n d qu a n t i ty o f p a d d y i na g ricult u re p r o du ct i o n . T heclassi f ic a tion o f p a dd yl e a f d iseaseisani m p o rta n ta n d u r g enttaskasit d estro y sa bou t 10 %to15% o f p rod u ct i on i n A si a . T h u s,a stu d y o na u to m a ticcla s s i f ic a t i o n o f p a d dyle a f d isease u si n gi m age p r o c e s s i n gis p resente d . F e a t u ree x tra c ti o nte c hn i q u es o fcol o r,t ex tur e ,a n d shape we re i m pl e m e ntedtoa n a l yz et h echara c teristics o ft h e p ad d yle a f d ise a s e .Inanother no te,aSuppo r t V ector M a c h i ne(S V M)is u sed tocla s sify t h e f o u rt y pes o f p a dd yle a f d is e ase w h icharet h e b r o w ns p o t, b a cteri a lle a f b light, t u n grovirus,a n d l eaf s c a l d . T he p er f o r m a n ce o ft h e p r opo s e dst ud yis e va luatedto 1 60testi n gi mage s w hichret u r n e d 86 .2 5 % o fcla s si f i c ation a cc ura c y . T he ou tc o m e o ft h isst ud yise x p e ctedtoassistt h ea g r o te c hno l o g y i n dustryine a r l y d ete c ti o n o f p a dd yle a f d iseasein w hichanap p ropriate a c ti o nc o uld b e ta ke n ac c or d in g l y .
TL;DR: In this article , the authors reviewed the latest low-illumination image enhancement methods based on deep learning and divided them into four categories: supervised learning, unsupervised learning, semi-supervised and zero-shot learning methods.
Abstract: As a critical preprocessing technique, low-illumination image enhancement has a wide range of practical applications. It aims to improve the visual perception of a given image captured without sufficient illumination. Conventional low-illumination image enhancement methods are typically implemented by improving image brightness, enhancing image contrast, and suppressing image noise simultaneously. Nevertheless, recent advances in this area are dominated by deep-learning-based solutions, and consequently, various deep neural networks have been proposed and applied to this field. Therefore, this paper briefly reviews the latest low-illumination image enhancement, ranging from its related algorithms to its unsolved open issues. Specifically, current low-illumination image enhancement methods based on deep learning are first sorted out and divided into four categories: supervised learning methods, unsupervised learning methods, semi-supervised learning methods, and zero-shot learning methods. Then, existing low-light image datasets are summarized and analyzed. In addition, various quality assessment indices for low-light image enhancement are introduced in detail. We also compare 14 representative algorithms in terms of both objective evaluation and subjective evaluation. Finally, the future development trend of low-illumination image enhancement and its open issues are summarized and prospected.
TL;DR: Wang et al. as mentioned in this paper proposed a 4D-NLM (non-local means) denoising method for 4D fetal heart ultrasound image sequence, which takes advantage of context similar information in neighboring images to denoise the target image, and an enhancing method called the Adaptive Clipping for Each Histogram Pillar (ACEHP), which is designed to enhance myocardial spaces to distinguish them from blood spaces.
Abstract: Background 4D ultrasound images of human fetal heart are important for medical applications such as evaluation of fetal heart function and early diagnosis of congenital heart diseases. However, due to the high noise and low contrast characteristics in fetal ultrasound images, denoising and enhancements are important. Methods In this paper, a special method framework for denoising and enhancing is proposed. It consists of a 4D-NLM (non-local means) denoising method for 4D fetal heart ultrasound image sequence, which takes advantage of context similar information in neighboring images to denoise the target image, and an enhancing method called the Adaptive Clipping for Each Histogram Pillar (ACEHP), which is designed to enhance myocardial spaces to distinguish them from blood spaces. Results Denoising and enhancing experiments show that 4D-NLM method has better denoising effect than several classical and state-of-the-art methods such as NLM and WNNM. Similarly, ACEHP method can keep noise level low while enhancing myocardial regions better than several classical and state-of-the-art methods such as CLAHE and SVDDWT. Furthermore, in the volume rendering after the combined "4D-NLM+ACEHP" processing, the cardiac lumen is clear and the boundary is neat. The Entropy value that can be achieved by our method framework (4D-NLM+ACEHP) is 4.84. Conclusions Our new framework can thus provide important improvements to clinical fetal heart ultrasound images.