Contrast Stretch Research Articles

A cross layer graphical neural network based convolutional neural network framework for image dehazing

The current version of imaging equipment cannot quickly and effectively make up for the reduction of visibility triggered by bad weather. Traditional strategies minimize hazy impacts by employing an image depth model and a physical model. Following experts, erroneous depth data reduces the efficacy of the dehazing algorithm. Dehazing methods based on CNN are imperfect to handle region which is bright or similar to atmospheric light and thus leads to oversaturation of pixels. These challenges can be addressed by proposing a novel model that incorporates the idea of a Graphical Neural Network. The amount of light coming from the atmosphere is estimated using normalization where the contrast of the image gets adjusted using Bias Contrast stretch Histogram Equalization. An enhanced Transmission map estimator is used to render the hazy scene. Finally, the cross-layer graphical neural network-based CNN model is applied to produce a haze-free image and eliminate the over-saturation of pixels. Extensive evaluation findings show that the proposed approach can significantly recuperate misty imagery, even if the images have a substantial amount of haze.

The Cuckoo Optimization Algorithm Enhanced Visualization of Morphological Features of Diabetic Retinopathy

This research compares strategies for identifying diabetic retinopathy (DR) using fundus image and discusses the efficiency of various image pre-processing techniques to enhance the quality of fundus images. Fundus images in medical image processing often suffer from non-uniform lighting, low contrast, and noise issues, which necessitate image pre-processing to enhance their quality. The study evaluates the effectiveness of several optimization techniques in selecting the best technique for identifying DR. One of the image pre-processing techniques compared in the study involves comparing negative images, dark contrast stretch, light contrast stretch, and partial contrast stretch, which are then evaluated using standard performance metrics such as NIQE, PNSR, MSE, and entropy. The results are further optimized using the Cuckoo Search Algorithm. The proposed technique produces better image quality improvements in several performance metrics, such as MSE, NIQE, PSNR, and entropy. Bright Contrast Stretch outperforms other techniques in NIQE Mean 5.2850, Entropy 5.0193, NIQE Standard deviation 0.2261, and Entropy 0.2612.

Retinex‐inspired contrast stretch and detail boosting for lowlight image enhancement

AbstractLowlight images with low brightness and contrast, blurry details usually bring us an uncomfortable visual experience. To promote the quality of these deviation images, this paper presents a new and efficient approach, named MFMR, for enhancing lowlight images in the hue‐saturation‐value (HSV) colour space. Concretely, the multi‐angle filter is first applied to estimate the artifact‐free illumination and reflection component of the V‐channel. Afterward, the adaptive bi‐interval histogram with human visual characteristics and morphological operations is employed to process the former, adaptive gamma correction to process the latter for generating various feature maps. In the end, these feature maps are united via adaptive multi‐scale fusion strategy to reconstruct high‐quality images, which are characterized by high contrast and brightness, vivid colour, and clearer details. Extensive experiments show that this method is a well‐proven low‐light image enhancement approach, which outperforms the state‐of‐the‐art comparison methods. Furthermore, the proposed method also can yield satisfying images in the heavy foggy, yellow sand, underwater, and other severe conditions.

Remodelling correlation: A fault resilient technique of correlation sensitive stochastic designs

Major sources of error in stochastic circuits are correlation errors, soft errors, and random fluctuation errors, which impact the circuit’s accuracy and reliability. The soft error has the effect of modifying the correlation status, which in turn modifies the probability of the output. This has serious implications for security and medical systems that require highly precise systems. To address this issue, we employ a fault-tolerant technique of correlation-sensitive stochastic logic circuits. To ensure reliable operation, we have developed a Remodelling Correlation ( ReCo ) framework for correlation-sensitive stochastic logic elements (SLEs). Using two intriguing case studies, we present two variants of ReCo model for combinational circuits with contradictory requirements. To achieve faster convergence to the desired Mean-Squared Error (MSE) value with less hardware area, the proposed method prioritizes the selection of logic elements and the placement of correction blocks. It is shown that the overall reliability of the circuit is unaffected by this method. To demonstrate the usefulness of the proposed framework, the contrast stretch operation on images of CEED2016, a standard contrast enhancement dataset, is investigated in a noisy setting. The average Multiscale structural similarity index (MS-SSIM) of the output images using the proposed method is observed as 0 . 91 ± 0 . 02 , which is significantly higher than the original images with error, i.e., 0 . 75 ± 0 . 12 .

Analyzing the Role of Emotional Intelligence on the Performance of Small and Medium Enterprises (SMEs) Using AI-Based Convolutional Neural Networks (CNNs)

Human emotion detection is necessary for social interaction and plays an important role in our daily lives. Artificial intelligence research is rising, focusing on automated emotion detection. The capability to identify the emotion, which is considered one of the traits of emotional intelligence, is a component of human intelligence. Although the study is limited dependent on facial expressions or voice is flourishing, it is identifying emotions via body movements, a less researched issue. To attain emotional intelligence, this study suggests a deep learning approach. Here initially the video can be converted into image frames after the converted image frames can be preprocessed using the Glitter bandpass butter worth filter and contrast stretch histogram equalization. Then from the enhanced image, the features can be clustered using the hybrid Gaussian BIRCH algorithm. Then the specialized features are retrieved from the body of human gestures using the AdaDelta bacteria foraging optimization algorithm, and the selected features are fed to a supervised Kernel Boosting LENET deep-learning algorithm. The experiment is conducted using Geneva multimodal emotion portrayals (GEMEPs) corpus data set. This data set includes, human body gestures portraying the archetypes of five emotions, such as anger, fear, joy, pride, and sad. In these emotion detection techniques, the suggested Kernel Boosting LENET classifier achieves 98.5% accuracy, 94% precision, 95% sensitivity, and F-Score 93% outperformed better than the other existing classifiers. As a result, emotional acknowledgment may help small and medium enterprises (SMEs) to improve their performance and entrepreneurial orientation. The correlation coefficient of 188 and the significance coefficient of 0.00 show that emotional intelligence and SMEs performance have a significant and positive association.

An Improved Hybrid Method for Defogging Single Image

In this article, we suggest a different method for addressing the issue of nighttime single image dehazing. Because a nighttime landscape frequently includes several light sources, ambient lighting for haze period is usually not globally isotropic. Existing nighttime dehazing algorithms have tried to treat these two zones using the same prior assumptions. We propose a novel blending approach for resolving them in this work. A channel difference guided filtering with contrast stretch approach is presented to estimate ambient light, which creates a spatially variable low-frequency passband that selectively retains high-frequency edge information.

Medical Images Enhanced by Using Fuzzy Logic Depending on Contrast Stretch Membership Function

Medical images are often adversely affected by a lack of clarity due to the limited representation of the color gamut. In this research, three types of medical images microscopic, magnetic resonance and x-ray images were enhanced by using a Fuzzy Logic by Stretch Membership Function (FLSMF). The Stretch Membership Function increased the dynamic range for the compounds red, green and blue in the medical images which have a few ranges. The FLSMF algorithm was compared with other methods by calculating the entropy value, wavelet quality evaluator and lightness order error. The analysis of the results showed that the proposed method succeeded in enhancing the contrast of the different types of medical images where it had high average values for the entropy (6.95) and wavelet quality evaluator (0.08), and a small average value for the lightness order error (60.77).

COMPARISON OF IMAGE ENHANCEMENT TECHNIQUES FOR RAPID PROCESSING OF POST FLOOD IMAGES

Abstract. Satellite images are widely used for assessing the areal extent of flooded areas. However, presence of clouds and shadow limit the utility of these images. Numerous digital algorithms are available for enhancing such images and highlighting areas of interest. These algorithms range from simple to complex, and the time required to process these images also varies considerably. For disaster response, it is important to select an algorithm that can enhance the quality of the images in relatively short time. This study compared the relative performance of five traditional (Histogram Equalization, Local Histogram Equalization, Contrast Limited Adaptive Histogram Equalization, Gamma Correction, and Linear Contrast Stretch) algorithms for enhancing post-flood satellite images. Flood images with different levels of clouds and shadows were enhanced and output generated were evaluated in terms of processing time and quality as measured by Blind/Reference less Image Spatial Quality Evaluator (BRISQUE), a no-reference image quality metric. Findings from this study will provide valuable information to image analysts for selecting a suitable algorithm for rapidly processing post-flood satellite images.

CT window trainable neural network for improving intracranial hemorrhage detection by combining multiple settings.

Window settings to rescale and contrast stretch raw data from radiographic images such as Computed Tomography (CT), X-ray and Magnetic Resonance images is a crucial step as data pre-processing to examine abnormalities and diagnose diseases. We propose a distant-supervised method for determining automatically the best window settings by attaching a window estimator module (WEM) to a deep convolutional neural network (DCNN)-based lesion classifier and training them in conjunction. Aside from predicting a flexible window setting for each raw image, we statistically identify the top four window settings by calculating the mean and standard deviations for the entire dataset. Images are scaled on each of the top settings estimated by WEM and following lesion classifiers are subsequently trained. We study the effects of only using the flexible window, the single fixed window as either a known default window used by radiologists or an estimated mean value, and two different approaches to combine results from the top window settings to improve the detection of intracranial hemorrhage (ICH) from brain CT images. Experimental results showed that using the top four window settings identified from the window estimator module and combining the results had the best performance.

An Adaptive Detail Equalization for Infrared Image Enhancement Based on Multi-Scale Convolution

In order to solve the problem of low contrast and fuzzy detail in infrared image, we propose an infrared image enhancement method based on multi-scale and adaptive bi-interval histogram equalization with details. The method mainly consists of four parts: details enhancement, contrast stretch, edge enhancement and reconstruction of enhancement images. Firstly, the multi-scale convolution is used to enhance the details of image; Secondly, taking maximize the variance between classes and minimize the variance as fitness function and solved the threshold of the infrared image by genetic algorithm, then dividing the infrared image into two sub-intervals according to the threshold. After that, the bi-interval histogram equalization with details is applied to enhance the global contrast, at the same time, using the mean square deviation and average gray equalization to improve the brightness of the image. Finally, the enhanced image by adaptive bi-interval histogram equalization with details and the image processed by adaptive limited Laplace operator are fused by linear weighting to reconstruct the final enhancement image. The experimental results show that the proposed method can outperform state-of-the-art ones in both qualitative and quantitative comparisons.

Deep learning based banana plant detection and counting using high-resolution red-green-blue (RGB) images collected from unmanned aerial vehicle (UAV).

The production of banana—one of the highly consumed fruits—is highly affected due to loss of certain number of banana plants in an early phase of vegetation. This affects the ability of farmers to forecast and estimate the production of banana. In this paper, we propose a deep learning (DL) based method to precisely detect and count banana plants on a farm exclusive of other plants, using high resolution RGB aerial images collected from Unmanned Aerial Vehicle (UAV). An attempt to detect the plants on the normal RGB images resulted less than 78.8% recall for our sample images of a commercial banana farm in Thailand. To improve this result, we use three image processing methods—Linear Contrast Stretch, Synthetic Color Transform and Triangular Greenness Index—to enhance the vegetative properties of orthomosaic, generating multiple variants of orthomosaic. Then we separately train a parameter-optimized Convolutional Neural Network (CNN) on manually interpreted banana plant samples seen on each image variants, to produce multiple results of detection on our region of interest. 96.4%, 85.1% and 75.8% of plants were correctly detected on three of our dataset collected from multiple altitude of 40, 50 and 60 meters, of same farm. Further discussion on results obtained from combination of multiple altitude variants are also discussed later in the research, in an attempt to find better altitude combination for data collection from UAV for the detection of banana plants. The results showed that merging the detection results of 40 and 50 meter dataset could detect the plants missed by each other, increasing recall upto 99%.

IRIS Recognition using Hough Transform

In most iris identification systems, the complete image acquires constraints are understood. These Constrain include near-infrared (NIR) illumination to release the iris texture and close distance from the capturing device. In recent advances to different illumination technologies introduced in images captured in the environment. This environment includes a visible wavelength (VW) light source at-a-distance over the close distance from the capturing device. For accurate Iris identification at-a-distance, eye images require improvement of effective strategies, while setting the light source at a distance from the planar view of the iris. Effectively performing feature extraction technique for Near-Infrared and Visible wavelength images, that were collected in an uncontrolled stage. The identification of iris accuracy on the publicly available databases was then measured. This paper presents a preprocessing of Iris Recognition using Hough Transform (HT) for Iris Area of interest (AOI) and rubber-sheeting the model captured using linear stretching and rotation for normalization. The HT is used to filter and contrast stretch the iris regions from multispectral iris images. A basic purpose of this research is to envelop a design and implement IRIS-recognition at a distance (IAAD) by adopting a frequency and wavelength-based Hough transform for accurate feature selection [1][2]. The proposed method is described as follows: Initially, the input iris image will be subjected to pre-processing while extracting features with differences from local extrema and maxima conditions, using a regular shape filling Hough transform [3][4]. The iris localization and detection consists of a hill climbing segmentation approach that is based on geometric shape Hough measure. Proposed in comparison to the contemporary

The potency of Sentinel-2 satellite for monitoring during and after coral bleaching events of 2016 in the some islands of Marine Recreation Park (TWP) of Pieh, West Sumatra

The coral bleaching event (CBE) which trigerred by increase in sea temperatures of 1-2˚C during El-Nino can make corals stressful or potentially dead. Many CBE research are conducted using sea temperature data measured through infrared thermal (TIR) satellite sensor. However, very few study use optical remote sensing data. This study aim to analyze the potential of optical Sentinel-2 (S-2) satellite data for monitoring during (Jun. 2016) and after (Jun. 2018) CBE in small islands of marine protected areas (TWP) of Pieh. A series image analysis procedure was conducted, i.e. simple DOS atmospheric correction; normalizating digital numbers (DN) of Blue, Green and Red bands of subject images during and after CBE to the reference image (before CBE. Dec. 2015) using “pseudo invariant feature (PIF)” approach; subtracting the DN of each bands of images during and after CBE to the reference image, and generating color composite of natural color with a contrast stretch. Result showed that this procedure can detected clearly the CBE. There was an average decreasing bleaching trend of 1.63 ha (15.8%), which indicated that some corals have recovered from thermal stress inside the TWP of Pieh. However, islands outside the TWP of Pieh, showed an increasing trend by an average of 0.50 ha (21.2%). The predation corals by the crown of thorn seastar, Acanthaster planci, whose population exploded after CBE was the main reason why CBE still detected both inside and outside the TWP of Pieh. This study shows the ability and potency of optical remote sensing data of S-2 satellite to monitor CBE, but need a lot of validation from many other sites. This result is very useful for supporting project such as COREMAP CTI.

Classification of Welding Flaws in Gamma Radiography Images Based on Multi-scale Wavelet Packet Feature Extraction Using Support Vector Machine

This paper has introduced a digital process aimed at automatically identify and classify flaws in the weld joints. Therefore, various algorithms are presented for computer aided detection (CAD) and classification. These algorithms include preprocessing algorithms using Gaussian pyramidal transform, contrast enhancement algorithm using contrast stretch and normalization method, noise reduction algorithm by blind image separation (BIS), measurement of the noise separation quality and image segmentation algorithm based on expectation-maximization (EM) method. Also, an algorithm for detection and classification of welding defects from radiographic images is presented. This algorithm is based on multi-scale wavelet packet (MWP) technique for feature extraction. Also, extraction of features from its transform domains is proposed to assist in achieving a higher classification rate. Moreover, the support vector machine (SVM) is applied for matching the extracted features. Consequently, classification error was computed for both normal and defect images. The obtained results confirm that much higher classification error was computed for defect image due to various objects may be in the image. Moreover, the accuracy of the considered algorithms is determined by statistical measurements. These algorithms have the potential for further improvement because of their simplicity and encouraging results. Therefore, it will motivate real-time flaws detection and classification for many CAD applications.

Digital canopy photography: Exposed and in the raw

Estimates of the canopy gap fraction, on which calculations of leaf area index (LAI) are based, are sensitive to photographic exposure in upward-facing images. In this article we describe a simple, automated method of image acquisition and processing that eliminates both subjectivity and the need for the operator to consider photographic exposure in the field. A key strength of our methodology was the use of a test apparatus (perforated aluminum screen) with a precisely known gap fraction; this allowed us to separate the confounding effects of gap size and gap fraction on the optimal photographic exposure for a canopy. We took photographs of the test apparatus at different photographic exposures; we varied the gap fraction by covering a proportion of the holes in the screen, and also varied gap size by varying the distance of the camera from the screen. We acquired both raw images and JPEG images. We found that the optimal exposure of JPEG images varied with both gap size and gap fraction, not just gap fraction as previously assumed. Underexposing by one stop yielded raw data that were never clipped resulting in data loss, but that used most of the 14-bit range of the raw file. We also found that it was easily possible to standardize photographic exposure during image processing by acquiring raw images in the field; thus eliminating the variation in estimated gap fraction and LAI associated with exposure variations. This result was replicated in both fisheye images and cover images that we acquired in real canopies. We recommend the following protocol for acquiring canopy images in future studies:1.Shoot raw using one stop of underexposure.2.Convert propriety-format raw files to DNG format using the free software Adobe DNG Converter.3.Contrast stretch the blue channel of the image and save as JPG for further analysis. The authors can supply MATLAB script to perform this step.

Linear and non-linear enhancement for sun glint reduction in advanced very high resolution radiometer (AVHRR) image

Cloud detection over water surfaces is difficult due to the sun glint effect. The mixed pixels between both features may introduce inaccurate cloud classification. This problem generally occurs because of less contrast between the glint and the cloud. Both features have almost the same reflectance in the visible wavelength. The piecewise contrast stretch technique shows preservation capability on the reflectance of the cloud. The result of a band ratio was smoothed by applying the Sobel edge detection to provide better cloud feature detection. The study achieved an accuracy of about 77.5% in cloud pixels detection.

ACCURATE AND FAST PUPIL LOCALIZATION USING CONTRAST STRETCHING, SEED FILLING AND CIRCULAR GEOMETRICAL CONSTRAINTS

Iris segmentation is the most contested issue in th e iris recognition system, since noise and poor ima ge quality can significantly affect accuracy of iris l ocalization stage. Therefore, very careful attentio n has to be paid for the segmentation process if only an accura te result is expected. This study presents a new me thod for precise pupil detection capable of handling the unconstrained bad acquisition conditions especiall y those related to low contrast or to the non-uniform brigh tness caused by the position of light sources, spec ular reflection, eyelashes and eyelids. Contrast stretch ing (normalization) technique is used for handling the variations in contrast and illumination in an iris image by stretching’ the range of intensity values. Next, the local integration is applied on the enhanced image, this process will enhance the contrast level betwe en the existing white and black areas of the image; this w ill useful to compute the optimal threshold value r equired to perform a successful image binarization for the purpose of isolation of the pupil region, the seed fill algorithm is used as region growing method to segment the binary image and allocate the pupil as a cir cular black segment with biggest area, the approximate pu pil center is detected then for removing the specul ar reflection, the pupil is filled with black color us ing a simple filling method. Finally a circle fitti ng algorithm is used for precisely allocating the circular pupil region by the fact that richer iris textures are n ot closer to the pupil boundary. A set of tests was conducted on 2,655 iris images which were downloaded from CASIA V3.0-interval standard dataset; the test results in dicated that the proposed method had subjectively 1 00% accuracy rate with pupil localization, process sati sfy the real time constraints even when dealing wit h images have very different brightness or contrast c onditions or they contain eyelashes artifacts.

Geological mapping of the Schuppen belt of north-east India using geospatial technology

A revised geologic map of the Schuppen belt of northeast India has been prepared based on interpretation of digitally enhanced satellite images. The satellite image interpretation is supported by limited field work and existing geologic maps. Available geological maps of this fold thrust belt are discontinuous and multi-scaled. The authors are of multiple opinions regarding the trajectory of formation boundaries and fault contacts. Digital image processing of satellite images and limited field surveys have been used to reinterpret and modify the existing geological maps of this fold thrust belt. Optical data of Landsat Thematic Mapper, Enhanced Thematic Mapper and elevation data of ASTER have been used to prepare this revised geological map. The study area extends from Hajadisa in south to Digboi oilfield in north, bounded by Naga thrust in the west and Disang thrust in the east. PCA, Image fusion, Linear Contrast stretch, Histogram Equalization and Painted relief algorithms have been used for the delineation of major geological lineaments like lithological boundary, thrust and strike slip faults. Digital elevation maps have enabled in the discrimination between thrust contacts and lithological boundaries, with the former being located mostly in the valleys. Textural enhancements of PCA, colour composites and Painted relief algorithm have been used to discriminate between different rock types. Few geological concepts about the terrain have been revisited and modified. It is assumed that this revised map should be of practical use as this terrain promises unexploited hydrocarbon reserves.

Brain MR image normalization in texture analysis of multiple sclerosis

A problem that occurs in quantitative texture analysis of magnetic resonance imaging (MRI) is that there are intra-scan and inter-scan image intensity variations due to the MRI instrumentation. Therefore, image intensity normalization methods should be applied prior to further image analysis. The objective of this work was to investigate six previously described MRI intensity normalization methods and propose the most appropriate for the pre-processing of brain T2-weighted MR images acquired from 38 symptomatic untreated multiple sclerosis (MS) subjects. The following normalization methods were investigated: Contrast Stretch Normalization (CSN), Intensity Scaling (IS), Histogram Stretching (HS), Histogram Normalization (HN), Gaussian Kernel Normalization (GKN), and Histogram Equalization (HE). The main findings of this study can be summarized as follows: 1) Lesion texture features were affected differently by the normalization process for both the 0 and 6-12 months MRI scans. For example, for the features median and contrast there was significant difference between 0 and 6-12 months for the original MRI images but not for the HN normalized ones. On the other hand for the feature complexity there was no significant difference between 0 and 6-12 months for the original MRI images, but there was for the HN normalized images. 2) The statistical lesion feature analysis between the original and the normalized images showed that the HN method gave the highest number of significant features after normalization for both the 0 and 6-12 months MRI scans. 3) The HN normalization method gave the best performance compared to the other normalization methods with respect to the distance measures, structural similarity index, coefficient of variation, and correlation coefficient between the original and the normalized 0 and 6-12 months MRI scans. Thus, based on the above findings it is recommended that the simple HN normalization method could be used prior to quantitative texture analysis in the case study presented. The findings of this study provide evidence that texture features of MRI-detectable brain white matter lesions may have an additional potential role in the clinical evaluation of MRI images in MS. This is a prerequisite step in the assessment of texture features as surrogate markers of disease progression. However, a larger scale study is needed to establish the application in clinical practice and for computing shape parameters and texture features that may provide information for better and earlier differentiation between normal tissue and MS lesions.

Quality Improvement for Analysis of Leukemia Images through Contrast Stretch Methods

Computer image processing develops very fast and plays an important role in wide applications of engineering and sciences. Leukemia is a malignant disease (Cancer) seen in people of any age groups either in children or adults aged over 50 years. In most of the cases microscopic images usually inadequate to identify the type of the cell. The traditional morphology test done by a hematologist to look under the microscope is a time consuming and tedious job. Diagnosis through this approach requires very costly equipment and may not be installed in all hospitals and clinics. Further the noises and blurriness effect during image acquisition often leads to false diagnosis of leukemia. An automatic image enhancement and segmentation system can make the inspection procedure of leukocytes much easier and faster and the amount of data that can be analyzed by such a clinician handle more data than they normally can handle. In this paper four contrast stretch based enhancement methods are implemented for analysis of leucocytes. A comparative analysis is done on local contrast stretching, global contrast stretching, dark contrast stretching, bright contrast stretching methods. All these methods involve threshold mapping which often useful to attain segmented results.