Identification Of Pixels Research Articles

Enhancement of Nighttime Fire Detection and Combustion Efficiency Characterization Using Suomi-NPP and NOAA-20 VIIRS Instruments

We present the second-generation Fire Light Detection Algorithm (FILDA-2), which includes advances in fire detection and retrievals of radiative power (FRP), fire visible energy fraction (VEF), and fire modified combustion efficiency (MCE) at nighttime from the holistic use of multiple-spectral radiances measured by the Visible Infrared Imaging Radiometer Suite (VIIRS) aboard Suomi-NPP (VNP) and NOAA-20/JPSS-1 (VJ1) satellites. Key enhancements include: (1) a new fast algorithm that maps VIIRS Day/Night Band (DNB) radiances to the pixel footprints of VIIRS Moderate (M) and Imagery (I) bands; (2) identification of potential fire pixels through the use of the DNB anomalies and I-Band thermal anomalies; (3) dynamic thresholds for contextual testing of fire pixels; and (4) pixel-specific estimates of FRP, VEF, and MCE. The global benchmark test demonstrates that FILDA-2 can detect approximately 25-30% more smaller and cooler fires than the operational VIIRS Active Fire 375 m I-band algorithm with the added benefit of providing daily global pixel-level characterizations of MCE for nighttime surface fires. The MCE derived by FILDA-2 is in good agreement with limited ground-based observations near the fires. Additionally, FILDA-2 reduces angular dependence in FRP estimates and significantly reduces the ”bow-tie” (double-counting) effect in fire detection compared to the AF-I product. The cross-validation of FILDA-2 products from VNP and VJ1 retrievals confirms good consistency in FRP and MCE retrievals globally. FILDA-2 is being implemented by NASA to generate a new VIIRS data product for fire monitoring, chemical-speciated fire emission estimates, and fire line characterization.

Despeckling Multitemporal Polarimetric SAR Data Based on Tensor Decomposition

Despeckling is an essential task in polarimetric synthetic aperture radar (PolSAR) image processing. Most existing filters developed for multi-temporal SAR images make use of either real or complex information. Real information refers to amplitude or intensity values, whereas complex information refers to complex covariance matrix (CCM) derived from either interferometric SAR (InSAR) or PolSAR data. The InSAR CCM is formed using images of the same polarimetric channel but acquired at different dates, and the PolSAR CCM contains information acquired simultaneously in different polarimetric channels. Therefore, these despeckling methods may present a good performance in some applications and scenes but fail in other cases, due to differences in input sources. In order to achieve a more robust result in all cases, we develop a method for multi-temporal polarimetric SAR data filtering based on tensor decomposition (TD-MPF), which aims at improving the identification of homogeneous pixels for spatially adaptive filtering. The key element of this approach consists of exploiting tensor theory to construct a new CCM that contains both polarimetric and interferometric information, as well as multi-temporal information for each pixel. The effectiveness of the proposed method and its performance are evaluated with simulated and real SAR data in comparison with several established methods.

Identification of bad pixels in shortwave infrared high-speed hyperspectral imager for spaceborne remote sensing

Identification of bad pixels in shortwave infrared high-speed hyperspectral imager for spaceborne remote sensing

New Real-Time Impulse Noise Removal Method Applied to Chest X-ray Images.

In this paper, we propose a new Modified Laplacian Vector Median Filter (MLVMF) for real-time denoising complex images corrupted by "salt and pepper" impulsive noise. The method consists of two rounds with three steps each: the first round starts with the identification of pixels that may be contaminated by noise using a Modified Laplacian Filter. Then, corrupted pixels pass a neighborhood-based validation test. Finally, the Vector Median Filter is used to replace noisy pixels. The MLVMF uses a 5 × 5 window to observe the intensity variations around each pixel of the image with a rotation step of π/8 while the classic Laplacian filters often use rotation steps of π/2 or π/4. We see better identification of noise-corrupted pixels thanks to this rotation step refinement. Despite this advantage, a high percentage of the impulsive noise may cause two or more corrupted pixels (with the same intensity) to collide, preventing the identification of noise-corrupted pixels. A second round is then necessary using a second set of filters, still based on the Laplacian operator, but allowing focusing only on the collision phenomenon. To validate our method, MLVMF is firstly tested on standard images, with a noise percentage varying from 3% to 30%. Obtained performances in terms of processing time, as well as image restoration quality through the PSNR (Peak Signal to Noise Ratio) and the NCD (Normalized Color Difference) metrics, are compared to the performances of VMF (Vector Median Filter), VMRHF (Vector Median-Rational Hybrid Filter), and MSMF (Modified Switching Median Filter). A second test is performed on several noisy chest x-ray images used in cardiovascular disease diagnosis as well as COVID-19 diagnosis. The proposed method shows a very good quality of restoration on this type of image, particularly when the percentage of noise is high. The MLVMF provides a high PSNR value of 5.5% and a low NCD value of 18.2%. Finally, an optimized Field-Programmable Gate Array (FPGA) design is proposed to implement the proposed method for real-time processing. The proposed hardware implementation allows an execution time equal to 9 ms per 256 × 256 color image.

An Improved Probability Density Function (PDF) for Face Skin Detection

Face Detection by skin color in the field of computer vision is a difficult challenge. Detection of human skin focuses on the identification of pixels and skin-colored areas of a given picture. Since skin colors are invariant in orientation and size and rapid to process, they are used in the identification of human skin. In addition features like ethnicity, sensor, optics and lighting conditions that are different are sensitive factors for the relationship between surface colors and lighting (an issue that is strongly related to color stability). This paper presents a new technique for face detection based on human skin. Three methods of Probability Density Function (PDF) were applied to detect the face by skin color; these are the Extreme Value Distribution Function and the Exponential Distribution Function methods, in addition to a new proposed model, over the HSV (Hue, Saturation, and Value) color space. The suggested technique aims to enhance skin pixel detection and improve the detection accuracy of a colored region in the human skin in a specific photo. The new model has proved to be 96.05% more accurate than the Extreme value distribution function and Exponential distribution function according to the selected region of the face during experiments. The images used in this paper were 380 color images from CalTech (California Technology Institute) dataset.

Distributed Scatterer Processing Based on Binary Partition Trees with Multi-Baseline PolInSAR Data

Distributed scatterers (DSs) are necessary to increase point density in multi-temporal InSAR (MT-InSAR) monitoring. The identification of homogeneous pixels (HPs) is the first and key step for DS processing to overcome the low signal-to-noise ratio condition. Since multi-polarization data are good at describing geometrical structures and dielectric properties of ground objects, they have been applied for HP identification. However, polarimetric information is not enough for identifying areas with similar ground objects but different deformation. We propose a novel DS preprocessing algorithm based on polarimetric interferometric homogeneous pixel (PIHP) identification. Firstly, a novel Polarimetric InSAR (PolInSAR) similarity that combines polarimetric intensity, interferometric coherence, and phase is proposed, which is readily available in multi-baseline and multi-polarization data and flexible by controlling weighting factors. Secondly, based on the binary partition tree (BPT) framework, object-orientated multi-scale PIHP identification is achieved, which is suitable for complex deformation scenes. Tested with simulated quad-polarization data, our method shows improvement in phase quality and point density, especially in the deformed areas, compared with the traditional HP identification method based on the polarimetric homogeneity (PolHom) test and the method with ground object type map. Tested with 30 quad-polarization Radarsat-2 images over Kilauea Volcano, the point density of our method is three times higher than that of the PolHom test in vegetation areas. Our method is proven to be more sensitive and mechanically more advanced to homogeneous pixels identification than the traditional ones, which is helpful for phase optimization, spatial enlargement of monitoring points, and stability of the MT-InSAR algorithm.

Demonstration of Neural Networks to Reconstruct Temperatures from Simulated Fluorescent Data Toward Use in Bio-microfluidics.

Biological systems often have a narrow temperature range of operation, which require highly accurate spatially resolved temperature measurements, often near ±0.1 K. However, many temperature sensors cannot meet both accuracy and spatial distribution requirements, often because their accuracy is limited by data fitting and temperature reconstruction models. Machine learning algorithms have the potential to meet this need, but their usage in generating spatial distributions of temperature is severely lacking in the literature. This work presents the first instance of using neural networks to process fluorescent images to map the spatial distribution of temperature. Three standard network architectures were investigated using non-spatially resolved fluorescent thermometry (simply-connected feed-forward network) or during image or pixel identification (U-net and convolutional neural network, CNN). Simulated fluorescent images based on experimental data were generated based on known temperature distributions where Gaussian white noise with a standard deviation of ±0.1 K was added. The poor results from these standard networks motivated the creation of what is termed a moving CNN, with an RMSE error of ±0.23 K, where the elements of the matrix represent the neighboring pixels. Finally, the performance of this MCNN is investigated when trained and applied to three distinctive temperature distributions characteristic within microfluidic devices, where the fluorescent image is simulated at either three or five different wavelengths. The results demonstrate that having a minimum of 10 3.5 data points per temperature and the broadest range of temperatures during training provides temperature predictions nearest to the true temperatures of the images, with a minimum RMSE of ±0.15 K. When compared to traditional curve fitting techniques, this work demonstrates that greater accuracy when spatially mapping temperature from fluorescent images can be achieved when using convolutional neural networks.

Reconstruction of land surface temperature under cloudy conditions from Landsat 8 data using annual temperature cycle model

Land surface temperature (LST) is an important parameter in the processes of energy exchange and water cycle between the land surface and the atmosphere. The impact of cloud cover leads to spatially incomplete of thermal infrared (TIR)-based LST products, which seriously hinders the applications of LST products in various fields. Several methods have been developed to reconstruct LST under cloudy conditions in previous studies, but there is a lack of an effective method for the reconstruction of cloudy LST at the spatial resolution of Landsat pixel (30 m). In this study, a novel method was proposed to reconstruct LST under cloudy conditions from Landsat 8 data. The LST reconstruction method includes four main steps: (1) identification of cloud-free, cloud-shadow, cloud-obscured, and cloud-covered pixels by integrating the Fmask method with a cloud-shape matching method; (2) calculation of annual temperature cycle (ATC)-based reference LST by fitting an ATC model to all available Landsat 8 LST product during 2013-2020; (3) estimation of LST residual from spatially adjacent similar pixels; and (4) estimation of reconstructed LST in terms of the sum of ATC-based reference LST and LST residual. The performance of the LST reconstruction method was evaluated using Landsat 8 LST images under clear-sky conditions as reference data. The root mean squared error (RMSE) between reconstructed LST and Landsat 8 reference LST ranges from 0.9 K to 2.5 K. The LST reconstruction method was further applied to reconstruct actual Landsat 8 LST images under cloudy conditions. Compared with original Landsat 8 LST images, the spatial distribution of reconstructed LST images is more complete. The pattern of reconstructed LST images reflects the spatial variability of LST well. The accuracy of the LST reconstruction method was validated against in situ LST measurements at six SURFRAD (Surface Radiation Budget Network) sites. The overall bias and RMSE between reconstructed LST and in situ LST at all sites are approximately −0.3 K and 3.5 K, respectively. The LST reconstruction method has great potentials to improve the applications of Landsat LST product in urban thermal environment monitoring and crop water stress monitoring.

RGB-Based Triple-Dual-Path Recurrent Network for Underwater Image Dehazing

In this paper, we present a powerful underwater image dehazing technique that exploits two image characteristics—RGB color channels and image features. In using RGB color channels, each color channel is decomposed into two units based on the similarities via the k-mean. This markedly improves the adaptability and identification of similar pixels, and thus reduces pixels with a weak correlation, leaving only pixels with a higher correlation. We use an infinite impulse response (IIR) in the triple-dual and parallel interaction structure to suppress hazed pixels via a pixel comparison and amplification to increase the visibility of even very minor features. This improves the visual perception of the final image, thus improving the overall usefulness and quality of the image. The softmax-weighted fusion is finally used to fuse the output color channel features to attain the final image. This preserves the color, leaving our proposed method’s output very true to the original scene’s. This is accomplished by taking advantage of adaptive learning based on the confidence levels of the pixel contribution variation in each color channel during subsequent fuses. The proposed technique both visually and objectively outperforms the existing methods in several rigorous tests.

Assessing Suitability of Auto-Selection of Hot and Cold Anchor Pixels of the UAS-METRIC Model for Developing Crop Water Use Maps

The METRIC energy balance model uses an auto-selection approach for identifying hot (dry, bare soil) and cold (fully transpiring crop) anchor pixels for the internal calibration of the model. When an unmanned aerial system (UAS) is used for imagery, the small image size and the varying crop and soil water status of agricultural fields make the identification of reliable hot and cold pixels challenging. In this study, we used an experimental spearmint field under three irrigation levels (75%, 100%, and 125% of crop evapotranspiration, ETc). As a way of providing diverse field conditions, six different extents (Extent 1 to Extent 6) were selected from each day of the seven days of UAS imagery campaigns of the same field for generating UAS-based ETc maps using auto-selection of hot and cold anchor pixels for the internal calibration of the model. Extent 1 had the smallest coverage area of the field, including only plants that were irrigated at 75% of ETc, while the fields of view of the other extents increased to where the Extent 6 covered the spearmint field and all the surroundings including trees, a nearby water canal, irrigated grass, and irrigated and non-irrigated soil. The results showed that different sizes of extent resulted in the selection of variable hot (bare, but moist soil in small extents, and dry bare soil at the larger extents) and cold anchor pixels (crop under water stress at the small extents, and tree canopy or grass alongside the water canal at the larger extents). This variation resulted in significantly different ETc estimation for the same spearmint crop field, indicative of a potential limitation for the use auto-selection of hot and cold pixels when using the UAS-METRIC model.

Automated Pixel-Level Deep Crack Segmentation on Historical Surfaces Using U-Net Models

Crack detection on historical surfaces is of significant importance for credible and reliable inspection in heritage structural health monitoring. Thus, several object detection deep learning models are utilized for crack detection. However, the majority of these models are powerful at most in achieving the task of classification, with primitive detection of the crack location. On the other hand, several state-of-the-art studies have proven that pixel-level crack segmentation can powerfully locate objects in images for more accurate and reasonable classification. In order to realize pixel-level deep crack segmentation in images of historical buildings, this paper proposes an automated deep crack segmentation approach designed based on an exhaustive investigation of several U-Net deep learning network architectures. The utilization of pixel-level crack segmentation with U-Net deep learning ensures the identification of pixels that are important for the decision of image classification. Moreover, the proposed approach employs the deep learned features extracted by the U-Net deep learning model to precisely describe crack characteristics for better pixel-level crack segmentation. A primary image dataset of various crack types and severity is collected from historical building surfaces and used for training and evaluating the performance of the proposed approach. Three variants of the U-Net convolutional network architecture are considered for the deep pixel-level segmentation of different types of cracks on historical surfaces. Promising results of the proposed approach using the U2−Net deep learning model are obtained, with a Dice score and mean Intersection over Union (mIoU) of 71.09% and 78.38% achieved, respectively, at the pixel level. Conclusively, the significance of this work is the investigation of the impact of utilizing pixel-level deep crack segmentation, supported by deep learned features, through adopting variants of the U-Net deep learning model for crack detection on historical surfaces.

Characterizing the calibration domain of remote sensing models using convex hulls

• We used a convex hull approach to assess model validity domain. • Convex hulls allow identifying unreliable model predictions. • At low sampling intensity, higher proportion of extrapolation are observed. • Prediction biases increase with extrapolation distances from the calibration domain. • This approach could help model development, as a complementary tool used in model quality assessment. The ever-increasing availability of remote sensing data allows production of forest attributes maps, which are usually made using model-based approaches. These map products are sensitive to various bias sources, including model extrapolation. To identify, over a case study forest, the proportion of extrapolated predictions, we used a convex hull method applied to the auxiliary data space of an airborne laser scanning (ALS) flight. The impact of different sampling efforts was also evaluated. This was done by iteratively thinning a set of 487 systematic plots using nested sub-grids allowing to divide the sample by two at each level. The analysis were conducted for all alternative samples and evaluated against 56 independent validation plots. Residuals of the extrapolated validation plots were computed and examined as a function of their distance to the model calibration domain. Extrapolation was also characterized for the pixels of the area of interest (AOI) to upscale at population level. Results showed that the proportion of extrapolated pixels greatly reduced with an increasing sampling effort. It reached a plateau (ca. 20% extrapolation) with a sampling intensity of ca. 250-calibration plots. This contrasts with results on model’s root mean squared error (RMSE), which reached a plateau at a much lower sampling intensity. This result emphasizes the fact that with a low sampling effort, extrapolation risk remains high, even at a relatively low RMSE. For all attributes examined (i.e. , stand density, basal area, and quadratic mean diameter) estimations were generally found to be biased for validation plots that were extrapolated. The method allows an easy identification of map pixels that are out of the calibration domain, making it an interesting tool to evaluate model transferability over an area of interest (AOI). It could also serve to compare “competing” models at a variable selection phase. From a model calibration perspective, it could serve a posteriori , to evaluate areas (in the auxiliary space) that merit further sampling efforts to improve model reliability.

Impact of bare soil pixels identification on clay content mapping using airborne hyperspectral AVIRIS-NG data: spectral indices versus spectral unmixing

Hyperspectral imaging spectroscopy has facilitated the mapping of soil properties at large scales, but since the presence of photosynthetic or non-photosynthetic vegetation affects the reflectance spectra, soil properties mapping is limited to bare soil surfaces. This study analyzed the impact of bare soil pixel identification on clay content estimation using two methods (i) combination of two spectral indices, Normalized Difference Vegetation Index for identifying photosynthetic vegetation and Cellulose Absorption Index for non-photosynthetic vegetation and (ii) spectral unmixing for estimating fractions of soil, photosynthetic and non-photosynthetic vegetation. The study used AVIRIS-NG image and laboratory measured clay content of 272 soil samples acquired over Karnataka, India. Bare soil pixels were identified using the two methods and performances of partial least squares regression (PLSR) models used to estimate the clay contents and the predicted clay content maps were analyzed and compared. PLSR model based on bare soil pixels identified by unmixing provided slightly better performances (R2 val of 0.61) than spectral indices (R2 val of 0.46), even though the percentage of study area mapped was reduced by half. This study highlighted that an improvement in prediction performance comes at the cost of reduction in spatial coverage in mapping of clay content.

Evaluation on sea surface wind speed product of Fengyun-3D satellite microwave radiation imager using ERA-5 dataset

ABSTRACT In this study, we evaluate the two years (from 1 May 2019 to 30 April 2021) global operational sea surface wind speed (SWS) product of Fengyun-3D (FY-3D) polar-orbit meteorological satellite microwave radiation imager (MWRI) using ERA-5 reanalysis data. The total mean bias error (MBE), mean absolute error (MAE) and correlation coefficient of FY-3D/MWRI SWS product are −0.432 ± 3.09 m/s, 1.896 ± 2.50 m/s and 0.76, respectively. The current FY-3D/MWRI SWS data are comparable with the wind speed data of ERA-5 at most of the sea areas, especially the farther offshore sea areas. However, we still find significant overestimations (more than 5 m/s) near the coasts and typical differences in horizontal MBE values in different ascending and descending orbits. Besides, the rainy pixels as another negative impact factor are also common in FY-3D/MWRI SWS product, which can induce manifest overestimations of SWS. Therefore, the future FY-3/MWRI SWS operational algorithm should further develop a rigorous and robust quality-control module and a joint rainy pixel identification algorithm to better eliminate the unreasonable (or anomalous) pixels and rainy pixels. Moreover, the detailed evaluation on the on-orbit calibration performance of MWRI sensor is also necessary for both ascending and descending orbits.

Classification and Identification of Spectral Pixels with Low Maritime Occupancy Using Unsupervised Machine Learning

For marine accidents, prompt actions to minimize the casualties and loss of property are crucial. Remote sensing using satellites or aircrafts enables effective monitoring over a large area. Hyperspectral remote sensing allows the acquisition of high-resolution spectral information. This technology detects target objects by analyzing the spectrum for each pixel. We present a clustering method of seawater and floating objects by analyzing aerial hyperspectral images. For clustering, unsupervised learning algorithms of K-means, Gaussian Mixture, and DBSCAN are used. The detection performance of those algorithms is expressed as the precision, recall, and F1 Score. In addition, this study presents a color mapping method that analyzes the detected small object using cosine similarity. This technology can minimize future casualties and property loss by enabling rapid aircraft and maritime search, ocean monitoring, and preparations against marine accidents.

UAV remote sensing based estimation of green cover during turfgrass establishment

Turfgrass is an important urban crop in the United States. Determining the percent green cover (PGC) to assess turfgrass quality/health and the rate of establishment is a crucial parameter for evaluating different species and experimental lines within species. However, evaluating the PGC of individual plots within large breeding nurseries in a conventional way, either visually or through digital image analysis is a time-consuming and laborious process. In the present study, we used the unmanned aerial vehicle (UAV) with multispectral and RGB sensors to estimate PGC during turfgrass establishment. We evaluated thirty approaches with different levels of complexity based on vegetation indices, supervised and unsupervised machine learning classification methods, and image processing methods for high-throughput turfgrass PGC estimation. An HSV (Hue-Saturation-Value) color space-based green pixel identification (GPI) method was introduced for the first time for estimating UAV derived PGC (UAVPGC). The results indicate that the GPI achieved the highest coefficient of determination, 0.86–0.96, with lowest mean absolute error when compared to ground percent green cover (GroundPGC). Overall, UAV-derived RGB image-based support vector machine methods were in agreement with GroundPGC (R2 = 0.88–0.95). This suggests that UAV-derived RGB images are adequate in accurately determining percent green cover (green vegetation within an experimental plot); however, multispectral images might offer a solution to determine turfgrass coverage (green and non-green vegetation within an experimental plot) during turfgrass establishment to account for non-green vegetation which is not captured by RGB (visible light spectrum) based estimation of PGC.

Pragmatic method for fast programming of hybrid photon counting detectors

It is now been over 15 years since Hybrid Photon Counting Detectors (HPCD) became one of the standard position-sensitive detectors for synchrotron light sources and X-ray detection applications. This is mainly due to their single-photon sensitivity over a high dynamic energy range and electronic noise suppression thanks to energy thresholding. To reach those performances, all HPCD pixels must feature the same electrical response against photons of the same energy. From the analysis of a monochromatic beam, in case of an ideal HPCD detector, it would be sufficient to apply a fixed voltage threshold among all pixels, positioned at half of the mean pulse amplitude to count every photon above the threshold. However, in practical cases, it must be considered that noise baselines from all pixels are not always strictly located at the same voltage level but can be spread over some voltage ranges. To address this kind of issue, most of all HPCDs apply a conventional threshold equalization method, that mainly relies on three steps; the setting of a global threshold at an arbitrary value, the identification of pixels noise baseline around that global threshold through an in-pixel threshold trimmer, and the computation of the required threshold offsets for setting all pixels at their own noise baseline at the same time. However, in case of a first-time use of an HPCD prototype, the threshold equalization might be biased by parameters that are wrongly set. Those biases can sometimes be characterized by the inability to localize some pixel noise baselines, which could be outside the voltage range of the threshold trimmer. The recovery of those biased pixels could be performed by changing the position of the global threshold, or by increasing the voltage range of the threshold trimmer. Unfortunately, both solutions could be time consuming due to the lack of information on the required steps for recovering all noise baselines. In order to overcome this issue in a reasonable time, this work introduces a pragmatic method that can be applied to HPCDs for an early and effective identification of appropriate pixels’ parameters, avoiding the need to test a high number of pixels configurations. The application of this method, at the early stage of the HPCD calibration, may drastically reduce the investigation time for finding the optimal operating parameters of HPCD prototypes.

Outlier-Robust Superpixel-Level CFAR Detector With Truncated Clutter for Single Look Complex SAR Images

The constant false alarm rate (CFAR) detectors are well studied for ship detection in SAR images, which suffer performance degradation due to the capture effect from interfering outliers, such as nearby targets, sidelobes and ghosts in multi-target environments. To address this issue, the clutter truncation scheme is adopted to reduce the outlier contamination in clutter samples such that the accuracy of clutter modeling can be improved. However, the selection of clutter truncation depth is difficult, which often resorts to sensitivity study. In this paper, the complex signal kurtosis (CSK) is first utilized as a statistical indicator for the decision of truncation depth to guarantee that the true clutter samples are maintained. Besides, a coarse-to-fine detection process is designed, including global superpixel proposal with the CSK and local identification of target pixels with the superpixel-level CFAR detector based on truncated statistics. During the local CFAR detection stage, the segmented superpixels provide convenient sample indexing for the iterative clutter truncation processing. The elevated performance achieved by the proposed method mainly benefits from the schemes of two-stage detection and automatic clutter truncation, yielding the increased detection efficiency and accuracy at the same time. Besides, false alarms caused by radio frequency interference can be reduced. In the experiment, the comparative results with state-of-the-art methods based on the Sentinel-1 and Gaofen-3 SAR data validate the performance of the proposed method.

Cusp Pixel Labelling Model for Objects Outline Using R-CNN

In order to accomplish machine visualization, object detection algorithms strive to recognise all target items in the image and derive the categories and object location information. Numerous ways to solve this problem have been presented where most of them are inspired by computer vision and deep learning methodologies. Existing techniques, on the other hand, consistently fail to recognise small, dense objects and even fail to detect objects with random geometric modifications. Object skeletons can help in object representation and detection. Skeleton is a natural object’s inherent visual description that contains comprehensive shape semantics. They supplement the object outline by providing additional information, such as how the object scale in thickness that varies between object components. However, extracting object skeletons from natural photos is difficult because the extractor must be able to capture both local and non-local visual contexts in order to calculate the scale of each skeleton pixel. To solve this issue, a Cusp Pixel Labelled Model with Precise Tuned Outline using Machine Learning (CPLM-PTOML) is proposed in this paper that accurately detects the cusp points of the object in the image by extracting the skeleton of the object to recognize the exact object resided in the image. A multi scale associated outputs to each stage of the image by monitoring the relationship between the receptive field sizes of the different layers in the image and the skeleton scales they can capture. The model is trained via multi-task learning, with one job being skeleton localization, which determines whether or not a pixel is a skeleton pixel, and the other being skeleton scale prediction, which predicts the scale of each skeleton pixel. At various phases, supervision is imposed by directing the scale-associated side outputs toward the ground-truth skeletons at the appropriate scales. The proposed model is compared with the traditional model in terms Data Training Samples, Object Feature Extraction Time Levels, Feature Extraction Accuracy, Cusp Pixel Labelling Time Levels, Cusp Pixel Identification Accuracy, Cusp Point Linking Accuracy, Image Considered and Cusp Point Recognition levels and the results exhibit that the proposed model exhibits better outcome.

Accurate estimation of camera response function for high dynamic range measurement.

A high dynamic range surface has large reflectivity variations and measuring it with a structured light technique could cause pixel saturation that will seriously degrade the measurement accuracy. Accurate identification of saturated pixels will significantly improve the measurement accuracy. This paper introduces an accurate estimation approach to the camera response function (CRF) for high dynamic range measurement and proposes a method that uses the estimated CRF to identify the saturated pixels in captured images. We perform experiments using the proposed saturation identification method and the other existing methods to compare both the saturation identification accuracy and measurement accuracy. The experiment results verify both the accuracy and suitability of the proposed method.