Single Grayscale Image Research Articles

Machine Fault Diagnosis through Vibration Analysis: Time Series Conversion to Grayscale and RGB Images for Recognition via Convolutional Neural Networks

Accurate and timely fault detection is crucial for ensuring the smooth operation and longevity of rotating machinery. This study explores the effectiveness of image-based approaches for machine fault diagnosis using data from a 6DOF IMU (Inertial Measurement Unit) sensor. Three novel methods are proposed. The IMU6DoF-Time2GrayscaleGrid-CNN method converts the time series sensor data into a single grayscale image, leveraging the efficiency of a grayscale representation and the power of convolutional neural networks (CNNs) for feature extraction. The IMU6DoF-Time2RGBbyType-CNN method utilizes RGB images. The IMU6DoF-Time2RGBbyAxis-CNN method employs an RGB image where each channel corresponds to a specific axis (X, Y, Z) of the sensor data. This axis-aligned representation potentially allows the CNN to learn the relationships between movements along different axes. The performance of all three methods is evaluated through extensive training and testing on a dataset containing various operational states (idle, normal, fault). All methods achieve high accuracy in classifying these states. While the grayscale method offers the fastest training convergence, the RGB-based methods might provide additional insights. The interpretability of the models is also explored using Grad-CAM visualizations. This research demonstrates the potential of image-based approaches with CNNs for robust and interpretable machine fault diagnosis using sensor data.

Highlight Removal from A Single Grayscale Image Using Attentive GAN

ABSTRACT The existence of specular highlights hinders high-level computer algorithms. In this paper, we propose a novel approach to remove specular highlights from a single grayscale image by regarding the problem as an image-to-image translation task between the highlight domain and the diffuse domain. We solve this problem by using the generative adversarial network framework, where a generator removes highlights and discriminator judges whether outputs of the generator are clear and highlight-free. Specular highlight removal is intractable as we should remove specular highlights while keeping as many details as possible. Considering the similarity between the highlight image and diffuse image, we adopt an attention-based submodule that generates a mask image, which we call the highlight intensity mask, to locate pixels that contain specular highlights and help the skip-connected autoencoder to remove highlights. A pixel discriminator and Structural Similarity loss are utilized to ensure that more details can be retained in the output images. For training and testing models, we build a grayscale highlight images dataset. It consists of more than a thousand sets of grayscale highlight images with ground truth. Finally, quantitative and qualitative evaluations demonstrate the effectiveness of our method than other contrast generative adversarial network methods.

A Zero-Watermark Scheme Based on Quaternion Generalized Fourier Descriptor for Multiple Images

Most of the existing zero-watermark schemes for medical images are only appropriate for a single grayscale image. When they protect a large number of medical images, repeating operations will cause a significant amount of time and storage costs. Hence, this paper proposes an efficient zero-watermark scheme for multiple color medical images based on quaternion generalized Fourier descriptor (QGFD). Firstly, QGFD is utilized to compute the feature invariants of each color image, then the representative features of each image are selected, stacked, and reshaped to generate a feature matrix, which is then binarized to get a binary feature image. Copyright information can be converted into the copyright image by using QR code technology, which contains more information. Finally, the zero-watermark image is constructed by executing the XOR operation on the copyright image and the feature image scrambled by the Cat map. In the experiment, different parameters are selected to determine the maximum number of images that the proposed scheme can protect simultaneously while achieving good robustness. The experimental results demonstrate that the proposed scheme can effectively resist common attacks, geometric attacks and joint attacks, and effectively improve the operation efficiency of the algorithm, thus effectively decreasing the time and storage cost of copyright protection for lots of medical images.

Neural Network-Based Pose Estimation for Noncooperative Spacecraft Rendezvous

This article presents the Spacecraft Pose Network (SPN), the first neural network-based method for on-board estimation of the pose, i.e., the relative position and attitude, of a known noncooperative spacecraft using monocular vision. In contrast to other state-of-the-art pose estimation approaches for spaceborne applications, the SPN method does not require the formulation of hand-engineered features and only requires a single grayscale image to determine the pose of the spacecraft relative to the camera. The SPN method uses a convolutional neural network (CNN) with three branches to solve the problem of relative attitude estimation. The first branch of the CNN bootstraps a state-of-the-art object detection algorithm to detect a 2-D bounding box around the target spacecraft in the input image. The region inside the 2-D bounding box is then used by the other two branches of the CNN to determine the relative attitude by initially classifying the input region into discrete coarse attitude labels before regressing to a finer estimate. The SPN method then estimates the relative position by using the constraints imposed by the detected 2-D bounding box and the estimated relative attitude. Further, with the detection of 2-D bounding boxes of subcomponents of the target spacecraft, the SPN method is easily generalizable to estimate the pose of multiple target geometries. Finally, to facilitate integration with navigation filters and perform continuous pose tracking, the SPN method estimates the uncertainty associated with the estimated pose. The secondary contribution of this article is the generation of the Spacecraft PosE Estimation Dataset (SPEED), which is used to train and evaluate the performance of the SPN method. SPEED consists of synthetic as well as actual camera images of a mock-up of the Tango spacecraft from the PRISMA mission. The synthetic images are created by fusing OpenGL-based renderings of the spacecraft's 3-D model with actual images of the Earth captured by the Himawari-8 meteorological satellite. The actual camera images are created using a seven degrees-of-freedom robotic arm, which positions and orients a vision-based sensor with respect to a full-scale mock-up of the Tango spacecraft with submillimeter and submillidegree accuracy. The SPN method, trained only on synthetic images, produces degree-level relative attitude error and cm-level relative position errors when evaluated on the actual camera images with a different distribution not used during training.

Steganography Technique for Embedding a Variety of Binary Images inside a Grayscale Image

Because of the rapid development and use of the Internet as a communication media emerged to need a high level of security during data transmission and one of these ways is "Steganography". This paper reviews the Least Signification Bit steganography used for embedding a variety of related binary images in a single grayscale image rather than traditional methods of embedding information which is embedded one image in another image (cover image). First three images represent the secret images which are used for hiding from unauthorized users while other images used for making more difficult and more secret watermark. In this paper, we have merged both methods of hiding and watermark. Instead of either creating a watermark or hiding secret data in a grayscale image, we can use both of them at the same image to take advantage of both the hiding and watermark properties at the same time. The findings of the research were the results as a new watermark composed of multiple related binary images and hidden secret multiple binary images inside.

Catheter segmentation in X-ray fluoroscopy using synthetic data and transfer learning with light U-nets

Background and objectivesAutomated segmentation and tracking of surgical instruments and catheters under X-ray fluoroscopy hold the potential for enhanced image guidance in catheter-based endovascular procedures. This article presents a novel method for real-time segmentation of catheters and guidewires in 2d X-ray images. We employ Convolutional Neural Networks (CNNs) and propose a transfer learning approach, using synthetic fluoroscopic images, to develop a lightweight version of the U-Net architecture. Our strategy, requiring a small amount of manually annotated data, streamlines the training process and results in a U-Net model, which achieves comparable performance to the state-of-the-art segmentation, with a decreased number of trainable parameters.MethodsThe proposed transfer learning approach exploits high-fidelity synthetic images generated from real fluroscopic backgrounds. We implement a two-stage process, initial end-to-end training and fine-tuning, to develop two versions of our model, using synthetic and phantom fluoroscopic images independently. A small number of manually annotated in-vivo images is employed to fine-tune the deepest 7 layers of the U-Net architecture, producing a network specialized for pixel-wise catheter/guidewire segmentation. The network takes as input a single grayscale image and outputs the segmentation result as a binary mask against the background.ResultsEvaluation is carried out with images from in-vivo fluoroscopic video sequences from six endovascular procedures, with different surgical setups. We validate the effectiveness of developing the U-Net models using synthetic data, in tests where fine-tuning and testing in-vivo takes place both by dividing data from all procedures into independent fine-tuning/testing subsets as well as by using different in-vivo sequences. Accurate catheter/guidewire segmentation (average Dice coefficient of ~ 0.55, ~ 0.26 and ~ 0.17) is obtained with both U-Net models. Compared to the state-of-the-art CNN models, the proposed U-Net achieves comparable performance ( ± 5% average Dice coefficients) in terms of segmentation accuracy, while yielding a 84% reduction of the testing time. This adds flexibility for real-time operation and makes our network adaptable to increased input resolution.ConclusionsThis work presents a new approach in the development of CNN models for pixel-wise segmentation of surgical catheters in X-ray fluoroscopy, exploiting synthetic images and transfer learning. Our methodology reduces the need for manually annotating large volumes of data for training. This represents an important advantage, given that manual pixel-wise annotations is a key bottleneck in developing CNN segmentation models. Combined with a simplified U-Net model, our work yields significant advantages compared to current state-of-the-art solutions.

Reducing overlapped pixels: a multi-objective color thresholding approach

This paper proposes a general multi-objective thresholding segmentation methodology for color images and a quality metric designed to prevent and quantify the overlapping effect of segmented images. Multi-level thresholding (MTH) has been used to segment color images in recent years; this process considers each channel as a single grayscale image and applies the MTH independently. Although this method provides competitive results, the inherent relationship among color channels is disregarded. Such approaches generate spurious classes on overlapping regions, where new colors are generated, especially on the borders of the objects. The proposed multi-objective color thresholding (MOCTH) approach performs image segmentation while preserving the relationship between image channels. MOCTH is aimed to reduce the overlapping effect on segmented color images without performing additional post-processing. To measure the overlapping classes on a thresholded color image, the overlapping index is proposed to quantify the pixels affected. The presented approach is analyzed on two color spaces (RGB and CIE L*a*b*) using three multi-objective algorithms; they are NSGA-III, SPEA-2, and MOPSO. Results provide evidence pointing out to a better segmentation from MOCTH over the traditional single-objective approaches while reducing overlapped areas on the image.

A Discriminative Colour-to-Grayscale Representation for Retinal Vessel Segmentation

Structural changes in retinal blood vessels are important indicators for many ocular diseases including diabetic retinopathy (DR), glaucoma, and hypertension. An automated vessel segmentation process almost always begins with acquired color fundus image, containing three layers of images (red, green and blue), and then quickly converted them to a single grayscale image. However, grayscale conversion is not unique and more than one grayscale representation can be obtained for a given color image. For the many currently existing automated vessel extraction methods, the green channel of the RGB color fundus image is routinely used as an input grey-scale representation to a pipeline of the segmentation process for the reason that it provides the best contrast among all three channels, namely red, green and blue. We hypothesize that vessel information contained in dropped channels, red and blue, will add to result in improved segmentation performance. In this paper, we propose a linear combination framework to utilize all three channels of the color fundus image based on their importance level rather than completely discarding some channels. We devised a Principal Component Analysis (PCA) method that provides appropriate weights for each color channel to realize a more discriminative grey-scale representation. The added information made available through PCA-based grayscale representation results in improved performance for Modautomated vessel segmentation algorithms. The performance of the framework is analyzed on two publically available databases (DRIVE, STARE) of fundus images quantifying improvements in all three aspects called accuracy, sensitivity, and specificity.

An RGB Image Cipher Using Chaotic Systems, 15-Puzzle Problem and DNA Computing

The application of confusion and diffusion processes on the three individual components of an RGB image is not secure and efficient, so this problem needs to be addressed. In this paper, a novel RGB image cipher is proposed using chaotic systems, 15-puzzle artificial intelligence problem and DNA computing. First of all the given color image is decomposed into its red, green and blue gray scale images. Then these gray scale images are concatenated to make a single gray scale image. This single gray scale image is further divided into different blocks. A block level permutation (BLP) is proposed on this gray scale image by using the 15-puzzle problem. A pixel level permutation is applied to further randomize the image pixels. This confused image is then DNA encoded. Afterwards, a diffusion process is applied on this DNA encoded image. Lastly this DNA diffused image is converted back into the decimal. Further, this single gray scale image is broken into three gray scale images. These three images are combined to get the final color cipher image. To create the plaintext sensitivity, SHA 256 hash function has been used. Both the simulation and a comprehensive security analyses suggest the robustness and the impregnability of the proposed scheme which in turn signals towards the real world applicability of the scheme.

Intrinsic Image Sequence Decomposition Using Low-Rank Sparse Model

Intrinsic image decomposition from a single image or image sequences is always a challenging task in computer vision and image processing due to the ill-posed characteristics. In order to attain a reasonable estimation of intrinsic images, in this paper, we present a low-rank sparse model (LRSM) to derive intrinsic images from an image sequence of the same scene under various lightings. Due to the dependance of varying lightings and the excellent edge-preserving ability of the total variation constraint, both are joined together to formulate the LRSM as a spatiotemporal prior. When considering the relationships among images of a scene as well as those between color channels of a color image, the proposed model involves a complex objective function and makes the solving much more difficult than that on a single gray-scale image. Therefore, we design an iterative numerical scheme based on the alternating direction method of multipliers framework to solve the objective function efficiently. We further specify the application of the proposed method to object recoloring. The experimental results demonstrate that the proposed LSRM and the iterative scheme are effective and efficient.

Multiple-image encryption method based on spatial angle multiplexing and double random phase encoding

An optical encryption method of multiple-image based on spatial angle multiplexing and double random phase encoding is proposed in this paper. In the encryption process, firstly the original images are modulated by random phase in Fresnel transform with different diffraction distances. Secondly, the modulated images are coherently superposed with reference beams which have different spatial angles and random phases, to generate interference fringes. Finally, the interference fringes from different directions are superposed to form a compound encrypted image. In the decryption process, the compound image is placed in a spatial filtering and Fresnel diffraction system, and the decrypted images are obtained after implementing the different phase keys’ demodulation and Fresnel diffraction with correct distance. This method encrypts multiple images into a single gray-scale image, which is easy to save and transmit. The double random phases are placed in object light and reference light respectively, which reduces the complexity of the encryption system and overcomes the difficulty of pixel-by-pixel alignment of random phase keys in traditional decryption experiment. At the same time, the multiplexing capacity of the proposed encryption system is analyzed, and the result shows that the system has sufficient encryption capacity. So the proposed method possesses the characteristics of high storage efficiency, simple calculation and strong anti-noise ability, and thus can encrypt multiple images simultaneously. In this paper, the encryption effect is evaluated by correlation coefficient, while the effectiveness and security are verified by simulation experiment.

Single-shot color image encryption based on mixed state diffractive imaging

We propose a method for color image encryption based on mixed state diffraction imaging. The three channels of the color image, R, G, B, are encrypted together into a single grayscale image by applying the classical optical double random phase mask encryption system, and mixed state decomposition algorithm is employed for the color image decryption, distinctively reducing the complexity of color image encryption with other methods which require to encrypt the color image into three separated channels. To the best of our knowledge, this is the first time that a color image is directly encrypted into a single grayscale encoded image. We have demonstrated the feasibility and high security of the proposed method with computer simulations.

Color-to-grayscale conversion through weighted multiresolution channel fusion

We present a color-to-gray conversion algorithm that retains both the overall appearance and the discriminability of details of the input color image. The algorithm employs a weighted pyramid image fusion scheme to blend the R, G, and B color channels of the input image into a single grayscale image. The use of simple visual quality metrics as weights in the fusion scheme serves to retain visual contrast from each of the input color channels. We demonstrate the effectiveness of the method by qualitative and quantitative comparison with several state-of-the-art methods.

A Specular Shape-from-Shading Method with Constrained Inpaiting

The existence of specular highlights is a great obstacle of shape-from-Shading. For a single gray-scale image with only intensity information, the existed highlights detection methods based on chroma or polarization analysis cannot directly be applied on it. So, a method using surface shape is provided, it makes full use of the imaging process. Through the surface normal estimation, compute the maximum a posteriori probability of each reflection composition under simulated annealing and detect the highlights areas. Finally, remove the highlights based on the assumption of curvature continuity. Experiments reveal that this method is effective on synthetic and real-world images, improves the accuracy of surface recovery for image combined specular highlights.

A Specular Highlight Removal Method with Shape Constrained Inpainting

The existence of specular highlight is a great obstacle of shape recovery. For a single gray-scale image with only intensity information, the existed highlight detection methods based on chroma or polarization analysis cannot directly be applied on it. So, a method using surface shape is provided, it makes full use of the imaging process. Through the surface normal estimation, compute the maximum a posteriori probability of each reflection composition under simulated annealing and detect the highlight areas. Finally, remove the highlight based on the assumption of curvature continuity. Experiments reveal that this method is effective on synthetic and real-world images, improves the accuracy of specular highlight removed for image combined specular highlight.

Solder Joints Detection Method Based on Surface Recovery

Machine vision has been widely used in various industrial productions. However, the study for solder joints detection is not enough. This paper presents a solder joints detection method based on surface recovery. For a single gray-scale image, using shape-from-shading (SFS) technology, the surface of the solder joints is recovered. According to the shape distribution, the quality of solder joints is discriminated. In order to improve the accuracy of recovery for real images, hybrid illumination model is introduced and a reflection-component estimation method based on simulated annealing algorithm is designed. Then recovery process of the algorithm is improved. Compared to other detection methods based on two-dimensional images, this method provides more information about explicit physical meaning and make detailed quantitative analysis for solder joints easier. At the same time, even for defect that is difficult to detect, this method also has important research value.

Estimating Facial Reflectance Properties Using Shape-from-Shading

In this paper we show how to estimate facial surface reflectance properties (a slice of the BRDF and the albedo) in conjunction with the facial shape from a single image. The key idea underpinning our approach is to iteratively interleave the two processes of estimating reflectance properties based on the current shape estimate and updating the shape estimate based on the current estimate of the reflectance function. For frontally illuminated faces, the reflectance properties can be described by a function of one variable which we estimate by fitting a curve to the scattered and noisy reflectance samples provided by the input image and estimated shape. For non-frontal illumination, we fit a smooth surface to the scattered 2D reflectance samples. We make use of a novel statistical face shape constraint which we term `model-based integrability' which we use to regularise the shape estimation. We show that the method is capable of recovering accurate shape and reflectance information from single grayscale or colour images using both synthetic and real world imagery. We use the estimated reflectance measurements to render synthetic images of the face in varying poses. To synthesise images under novel illumination, we show how to fit a parametric model of reflectance to the estimated reflectance function.

Shape from shading based on needle map and cellular automata

This paper presents a method for computing depth from a single grayscale image in two steps. Firstly, surface normals are parallelly and gradually adjusted by a procedure which includes three constraints: the smooth constraint ensures the recovered normals are smooth and integrable, the intensity gradient constraint ensures the recovered normals are consistent with the image gradient field, and the intensity constraint guarantees the recovered intensity is equal to the input image. Unlike their usage in global methods, those constraints are separately used in local area in our method. Secondly, the surface is recovered from needle map using a two-dimensional cellular automata system. An experimental assessment is provided for our methods on both real world images and synthetic images with known ground truth. The experiment results demonstrate this approach is practicable and has better precision than traditional methods.

Detecting textured objects using convex hull

In this paper, we present a methodology of locating 3D objects of known shapes from a single gray-scale image, in particular objects with rich textures on the surface. While traditional approaches identify objects by grouping and matching local features, we locate the object in the image using its convex hull, a high level feature not given much attention in the image using literature. A “direct line detection” algorithm is developed to detect line segments directly from the gray-scale image divided in small blocks. Lines are clustered and convex hull of a single or group of clusters is computed and edited to extract the 2D contour of the object. Successful experiments on rectangular boxes and cylinders show the effectiveness of the convex hull approach and its potential usage in industrial applications.

Segmentation of multi-dimensional infrared imagery from histograms

We present a technique for segmenting multi-dimensional data cubes based on multi-dimensional histograms. The histograms are formed from single gray-scale image reductions of the data cube such as principal component images. A segmentation is effected by associating each pixel with one of the peaks in the histogram. No spatial constraints are imposed and no training pixels are required. The following refinements to this simple process are described: proper weighting of the different principal components as a function of the peak shape; and automatic methods based on an entropy measure to generate a reasonable segmentation at a specified number of levels. Examples from both visible and infrared hyperspectral data will be shown.