Image Size Reduction Research Articles

Parallel Lossless Compression of Raw Bayer Images on FPGA-Based High-Speed Camera

Digital image compression is applied to reduce camera bandwidth and storage requirements, but real-time lossless compression on a high-speed high-resolution camera is a challenging task. The article presents hardware implementation of a Bayer colour filter array lossless image compression algorithm on an FPGA-based camera. The compression algorithm reduces colour and spatial redundancy and employs Golomb–Rice entropy coding. A rule limiting the maximum code length is introduced for the edge cases. The proposed algorithm is based on integer operators for efficient hardware implementation. The algorithm is first verified as a C++ model and later implemented on AMD-Xilinx Zynq UltraScale+ device using VHDL. An effective tree-like pipeline structure is proposed to concatenate codes of compressed pixel data to generate a bitstream representing data of 16 parallel pixels. The proposed parallel compression achieves up to 56% reduction in image size for high-resolution images. Pipelined implementation without any state machine ensures operating frequencies up to 320 MHz. Parallelised operation on 16 pixels effectively increases data throughput to 40 Gbit/s while keeping the total memory requirements low due to real-time processing.

Unishyper: A Rust-based unikernel enhancing reliability and efficiency of embedded systems

Unikernels are simple, customizable, efficient, and small in code size, which makes them highly applicable to embedded scenarios. However, most existing unikernels are developed and optimized for cloud computing, and they do not fully meet the requirements of high reliability and platform customization in embedded environments. We propose Unishyper, a reliable and high-performance embedded unikernel in Rust. To support memory isolation between user applications, user code, and kernel code, Unishyper designs the Zone mechanism on top of Intel MPK. Unishyper further proposes a thread-level unwind strategy for safe fault handling while avoiding memory leakage. Finally, Unishyper supports fine-grained customization, seamlessly integrates with the Rust ecosystem, and uses Unilib for function offloading to further reduce image size. Our evaluation results show that Unishyper achieves better performance than peer unikernels on major micro-benchmarks, can effectively stop illegal memory accesses across application boundaries, and has a minimal memory footprint of less than 100 KB.

Cross-patch feature interactive net with edge refinement for retinal vessel segmentation

Retinal vessel segmentation based on deep learning is an important auxiliary method for assisting clinical doctors in diagnosing retinal diseases. However, existing methods often produce mis-segmentation when dealing with low contrast images and thin blood vessels, which affects the continuity and integrity of the vessel skeleton. In addition, existing deep learning methods tend to lose a lot of detailed information during training, which affects the accuracy of segmentation. To address these issues, we propose a novel dual-decoder based Cross-patch Feature Interactive Net with Edge Refinement (CFI-Net) for end-to-end retinal vessel segmentation. In the encoder part, a joint refinement down-sampling method (JRDM) is proposed to compress feature information in the process of reducing image size, so as to reduce the loss of thin vessels and vessel edge information during the encoding process. In the decoder part, we adopt a dual-path model based on edge detection, and propose a Cross-patch Interactive Attention Mechanism (CIAM) in the main path to enhancing multi-scale spatial channel features and transferring cross-spatial information. Consequently, it improve the network’s ability to segment complete and continuous vessel skeletons, reducing vessel segmentation fractures. Finally, the Adaptive Spatial Context Guide Method (ASCGM) is proposed to fuse the prediction results of the two decoder paths, which enhances segmentation details while removing part of the background noise. We evaluated our model on two retinal image datasets and one coronary angiography dataset, achieving outstanding performance in segmentation comprehensive assessment metrics such as AUC and CAL. The experimental results showed that the proposed CFI-Net has superior segmentation performance compared with other existing methods, especially for thin vessels and vessel edges. The code is available at https://github.com/kita0420/CFI-Net.

Object-Oriented and Visual-Based Localization in Urban Environments.

In visual-based localization, prior research falls short in addressing challenges for the Internet of Things with limited computational resources. The dominant state-of-the-art models are based on separate feature extractors and descriptors without consideration of the constraints of small hardware, the issue of inconsistent image scale, or the presence of multi-objects. We introduce "OOPose", a real-time object-oriented pose estimation framework that leverages dense features from off-the-shelf object detection neural networks. It balances between pixel-matching accuracy and processing speed, enhancing overall performance. When input images share a comparable set of features, their matching accuracy is substantially heightened, while the reduction in image size facilitates faster processing but may compromise accuracy. OOPose resizes both the original library and cropped query object images to a width of 416 pixels. This adjustment results in a 2.4-fold improvement in pose accuracy and an 8.6-fold increase in processing speed. Moreover, OOPose eliminates the need for traditional sparse point extraction and description processes by capitalizing on dense network backbone features and selecting the detected query objects and sources of object library images, ensuring not only 1.3 times more accurate results but also three times greater stability compared to real-time sparse ORB matching algorithms. Beyond enhancements, we demonstrated the feasibility of OOPose in an autonomous mobile robot, enabling self-localization with a single camera at 10 FPS on a single CPU. It proves the cost-effectiveness and real-world applicability of OOPose for small embedded devices, setting the stage for potential markets and providing end-users with distinct advantages.

Role of Compressed Medical Images In Healthcare

This paper will focus on techniques used in patients monitoring in healthcare based on medical images. Lossless coding algorithm, watermark compression (add safety to data), GLCM, Huffman coding, LZW, DCT and soft compression techniques are used for efficient image compression. Lossless coding algorithms helps to reduce image size without losing the data. An image watermarking secret key (ROI) is used in this research to create a watermark that is unique to each picture. By comparing LZW compression with traditional approaches, it was observed that the LZW methodology has better performance than others.

Streamlined approaches for image classification using principal component analysis and hierarchical clustering of extrudates from coffee and sorghum blends

ABSTRACT This article describes simple methods to group images including principal component analysis (PCA) and hierarchical clustering of principal components (HCPC). Images of expanded and low expanded extrudates were processed using two optimization alternatives: a) image size reduction (from 2126 to 25 pixels); and b) grayscale conversion before size reduction. After applying PCA and HCPC, all tests yielded consistently similar results with the same PCA distribution and identical HCPC groups. Furthermore, expanded and low expanded extrudates formed groups with their respective peers. The RAM allocated to images and the time required to process them was reduced from 1727 Mb to less than 5 Mb and from ~ 2000s to just 0.1s, respectively. These results demonstrate the feasibility of using these two simple multivariate statistical techniques for image classification.

No threat: Emotion regulation neurofeedback for police special forces recruits

Police officers of the Special Forces are confronted with highly demanding situations in terms of stress, high tension and threats to their lives. Their tasks are specifically high-risk operations, such as arrests of armed suspects and anti-terror interventions. Improving the emotion regulation skills of police officers might be a vital investment, supporting them to stay calm and focused. A promising approach is training emotion regulation by using real-time (rt-) fMRI neurofeedback. Specifically, downregulating activity in key areas of the fronto-limbic emotion regulation network in the presence of threatening stimuli.Thirteen recruits of the Dutch police special forces underwent six weekly rt-fMRI sessions, receiving neurofeedback from individualized regions of their emotion regulation network. Their task was to reduce the image size of threatening images, wherein the image size represented their brain activity. A reduction in image size represented successful downregulation. Participants were free to use their preferred regulation strategy. A control group of fifteen recruits received no neurofeedback. Both groups completed behavioural tests (image rating on evoked valence and arousal, questionnaire) before and after the neurofeedback training. We hypothesized that the neurofeedback group would improve in downregulation and would score better than the control group on the behavioural tests after the neurofeedback training.Neurofeedback training resulted in a significant decrease in image size (t(12) = 2.82, p = .015) and a trend towards decreased activation in the target regions (t(10) = 1.82, p = .099) from the first to the last session. Notably, subjects achieved downregulation below the pre-stimulus baseline in the last two sessions. No relevant differences between groups were found in the behavioural tasks.Through the training of rt-fMRI neurofeedback, participants learned to downregulate the activity in individualized areas of the emotion regulation network, by using their own preferred strategies. The lack of behavioural between-group differences may be explained by floor effects. Tasks that are close to real-life situations may be needed to uncover behavioural correlates of this emotion regulation training.

SPA-UNet: A liver tumor segmentation network based on fused multi-scale features.

Liver tumor segmentation is a critical part in the diagnosis and treatment of liver cancer. While U-shaped convolutional neural networks (UNets) have made significant strides in medical image segmentation, challenges remain in accurately segmenting tumor boundaries and detecting small tumors, resulting in low segmentation accuracy. To improve the segmentation accuracy of liver tumors, this work proposes space pyramid attention (SPA)-UNet, a novel image segmentation network with an encoder-decoder architecture. SPA-UNet consists of four modules: (1) Spatial pyramid convolution block (SPCB), extracting multi-scale features by fusing three sets of dilated convolutions with different rates. (2) Spatial pyramid pooling block (SPPB), performing downsampling to reduce image size. (3) Upsample module, integrating dense positional and semantic information. (4) Residual attention block (RA-Block), enabling precise tumor localization. The encoder incorporates 5 SPCBs and 4 SPPBs to capture contextual information. The decoder consists of the Upsample module and RA-Block, and finally a segmentation head outputs segmented images of liver and liver tumor. Experiments using the liver tumor segmentation dataset demonstrate that SPA-UNet surpasses the traditional UNet model, achieving a 1.0 and 2.0% improvement in intersection over union indicators for liver and tumors, respectively, along with increased recall rates by 1.2 and 1.8%. These advancements provide a dependable foundation for liver cancer diagnosis and treatment.

Optimizing Storage and Computational Efficiency: An Efficient Algorithm for Whole Slide Image Size Reduction

ObjectiveTo efficiently store, transfer, and analyze whole slide imaging (WSI), we developed an image-processing algorithm to remove the unneeded background in a WSI and assemble tissue-containing parts into smaller WSIs without any change in tissue area image resolution. Patients and MethodsWe used histology slides of nondysplastic Barrett esophagus, low-grade dysplasia, and high-grade dysplasia, which were digitized using Aperio AT2 Scanner from January 1992 to September 2020. The algorithm involved converting color images to grayscale images, binarizing images by assigning zero to the background and 1 to the foreground, filling the holes and dilating the foreground masks, and extracting connected components. Using the coordinates of each component, the vertices of the smallest surrounding bounding box were calculated, and tissue-containing parts were cropped from the original slide. The smallest possible rectangle that encloses all bounding boxes containing tissue was found using the rectangle-packer package. ResultsThe algorithm resulted in a mean reduction of 7.11×. The performance of a previously developed deep learning model for the detection of Barrett esophagus dysplasia grade on the size-reduced WSIs was comparable with that on the original WSIs. ConclusionOur algorithm for WSI size reduction can assist researchers in storing, transferring, and analyzing WSIs while optimally using them in their workflow.

LAPLACIAN-BASED BLUR DETECTION ALGORITHM FOR DIGITAL BREAST TOMOSYNTHESIS IMAGES IN IMPROVING BREAST CANCER DETECTION

The most challenging aspect of working with digital images captured in an uncontrolled environment is to determine whether the image is of sufficient quality to be studied further. One of the most frequent reasons for a decrease in the quality of digital images is the presence of blur artefacts, especially when the images are taken from various angles of the x-ray source with limited angular range such as in digital breast tomosynthesis (DBT). The unwanted artefacts might substantially obscure the breast cancer location, especially in extremely dense fibroglandular breast tissue. It is almost impossible to differentiate breast cancer lesions in blurry and lowcontrast DBT images, thereby reducing the accuracy of lesion diagnosis. Due to this blurry artefact issue, this study aims to assess the performance of Laplacian-based Blur Detection (LbBD) algorithm for the blurry detection of DBT images. The LbBD algorithm is designed and developed using MATLAB R2021a software. The algorithm identifies the amount of blurriness by calculating the image variance; the farther the variance is from the threshold value, the less blurry is the image. On the contrary, the lower the variance level is from the threshold value, the greater the blur level. An online survey was conducted with an expert to assess the quality of 20 DBT images using two subjective measurements of blur or non-blur to ascertain the algorithm's performance. Three threshold values were used to compare the outcome with the output of the algorithm's status. With a low error rate (0.05), an accuracy of 95% at the ideal threshold value of 200, and an image size reduction of 10%, the system successfully predicts blurry images.

LWDS: lightweight DeepSeagrass technique for classifying seagrass from underwater images

In many coastal areas around the world, the seagrasses provide an essential source of livelihood for many civilizations and support high levels of biodiversity. Seagrasses are highly valuable, as they provide habitat for numerous fish, endangered sea cows, Dugong dugon, and sea turtles. The health of seagrasses is being threatened by many human activities. The process of seagrass conservation requires the annotation of every seagrass species within the seagrass family. The manual annotation procedure is time-consuming and lacks objectivity and uniformity. Automatic annotation based on lightweight DeepSeagrass (LWDS) is proposed to solve this problem. LWDS computes combinations of various resized input images and various neural network structures, to determine the ideal reduced image size and neural network structure with satisfactory accuracy and within a reasonable computation time. The main advantage of this LWDS is it classifies the seagrasses quickly and with lesser parameters. The DeepSeagrass dataset is used to test LWDS's applicability.

Wave Detection and Tracking Within a Rotating Detonation Engine Through Object Detection

As the operational time window of experimental rotating detonation engines (RDEs) is expanded and the technology matures toward integration within gas turbines, monitoring techniques must evolve to offer computationally efficient and highly time-resolved diagnostics. A computer vision object detection methodology that seeks to reduce data processing time and calculate wave velocity within drastically reduced time intervals as compared to traditional high-frame-rate RDE images analysis techniques is proposed. The adapted you-only-look-once object detection network is trained to detect individual detonation waves within single down-axis RDE images. The wave location and rotational direction detected within a frame are tracked through a series of high-speed images to calculate the frame-to-frame wave velocity with the time-step resolution of across a series of frames. The analysis of the annotation box size and image linearization effects is presented, demonstrating the lowest frame-to-frame velocity total uncertainty of and the highest classification speed of 9.5 frames per second using linearized images. Linearized images “unwrap” the RDE annulus pixel region to a reduced image size. This new method offers great reductions in data processing times and unsteady detonation behavior insight at intervals more comparable to the timescales of detonation wave interactions via the application of machine learning to experimental RDE data.

Super-resolution and segmentation deep learning for breast cancer histopathology image analysis.

Traditionally, a high-performance microscope with a large numerical aperture is required to acquire high-resolution images. However, the images' size is typically tremendous. Therefore, they are not conveniently managed and transferred across a computer network or stored in a limited computer storage system. As a result, image compression is commonly used to reduce image size resulting in poor image resolution. Here, we demonstrate custom convolution neural networks (CNNs) for both super-resolution image enhancement from low-resolution images and characterization of both cells and nuclei from hematoxylin and eosin (H&E) stained breast cancer histopathological images by using a combination of generator and discriminator networks so-called super-resolution generative adversarial network-based on aggregated residual transformation (SRGAN-ResNeXt) to facilitate cancer diagnosis in low resource settings. The results provide high enhancement in image quality where the peak signal-to-noise ratio and structural similarity of our network results are over 30 dB and 0.93, respectively. The derived performance is superior to the results obtained from both the bicubic interpolation and the well-known SRGAN deep-learning methods. In addition, another custom CNN is used to perform image segmentation from the generated high-resolution breast cancer images derived with our model with an average Intersection over Union of 0.869 and an average dice similarity coefficient of 0.893 for the H&E image segmentation results. Finally, we propose the jointly trained SRGAN-ResNeXt and Inception U-net Models, which applied the weights from the individually trained SRGAN-ResNeXt and inception U-net models as the pre-trained weights for transfer learning. The jointly trained model's results are progressively improved and promising. We anticipate these custom CNNs can help resolve the inaccessibility of advanced microscopes or whole slide imaging (WSI) systems to acquire high-resolution images from low-performance microscopes located in remote-constraint settings.

ECiS: Encryption prior to compression for digital image security with reduced memory

Digital images are a popular source of information and play an important role in several applications. However, the large amount of available data in the form of images causes serious security concerns. Although digital images are widely available, the storage of these images requires a large amount of data. Compression is an efficient way to reduce image size and efficiently utilize network resources. In this study, a joint encryption and compression technique, namely, ECiS, is developed to solve the two major issues with digital images outlined above. To achieve high security, we encrypt an image by using DNA, SHA-256, and a chaotic-based encryption technique. To reduce the bandwidth or storage space demand, we compress the encrypted image by using a compression technique based on zero memory set partitioned embedded block (ZM-SPECK). The strength of our ECiS technique is performing the analysis by applying several standard tests, including key analysis, statistical analysis, differential analysis, and time cost evaluation on a USC-SIPI dataset. Furthermore, the extensive evaluations on the dataset demonstrate that the proposed technique is secure and has a lower memory and encryption overhead than similar techniques.

Compression of Medical Images Using Wavelet Transform and Metaheuristic Algorithm for Telemedicine Applications

Medical image compression becomes necessary to efficiently handle huge number of medical images for storage and transmission purposes. Wavelet transform is one of the popular techniques widely used for medical image compression. However, these methods have some limitations like discontinuity which occurs when reducing image size employing thresholding method. To overcome this, optimization method is considered with the available compression methods. In this paper, a method is proposed for efficient compression of medical images based on integer wavelet transform and modified grasshopper optimization algorithm. Medical images are pre-processed using hybrid median filter to discard noise and then decomposed using integer wavelet transform. The proposed method employed modified grasshopper optimization algorithm to select the optimal coefficients for efficient compression and decompression. Four different imaging techniques, particularly magnetic resonance imaging, computed tomography, ultrasound, and X-ray, were used in a series of tests. The suggested method's compressing performance is proven by comparing it to well-known approaches in terms of mean square error, peak signal to noise ratio, and mean structural similarity index at various compression ratios. The findings showed that the proposed approach provided effective compression with high decompression image quality.

Three-Dimensional Adaptive Image Compression Concept for Medical Imaging: Application to Computed Tomography Angiography for Peripheral Arteries.

Advances in computed tomography (CT) have resulted in a substantial increase in the size of datasets. We built a new concept of medical image compression that provides the best compromise between compression rate and image quality. The method is based on multiple contexts and regions-of-interest (ROI) defined according to the degree of clinical interest. High priority areas (primary ROIs) are assigned a lossless compression. Other areas (secondary ROIs and background) are compressed with moderate or heavy losses. The method is applied to a whole dataset of CT angiography (CTA) of the lower extremity vasculature. It is compared to standard lossy compression techniques in terms of quantitative and qualitative image quality. It is also compared to standard lossless compression techniques in terms of image size reduction and compression ratio. The proposed compression method met quantitative criteria for high-quality encoding. It obtained the highest qualitative image quality rating score, with a statistically significant difference compared to other methods. The average compressed image size was up to 61% lower compared to standard compression techniques, with a 9:1 compression ratio compared with original non-compressed images. Our new adaptive 3D compression method for CT images can save data storage space while preserving clinically relevant information.

Classification Prediction Model Based on BP Neural Network

In this paper, a method of plant image processing combined with BP neural network was proposed to predict the disease degree of brassica napus. First, the initial image preprocessing, the roots of brassica napus. were used by Otsu threshold segmentation algorithm is combined with morphological open operation, segment the rape root image with better effect, in order to reduce image size inconsistent on the effects of extracting feature parameters, can be reconstructed images after segmentation, and then obtain the root image multiple characteristic parameters, Through chart correlation analysis to select the feature parameter information, with the characteristics of the rape root disease was positively associated with the degree of disease information to refactor pixel area and the longest diameter, finally establish the BP neural network, two characteristic parameters extracted as input, root disease degree level as the output, training network and prediction and comparison. The experimental results showed that the mean square error of the classification network for brassica olerica was 0.049, and the accuracy rate of the classification results predicted by the model was about 93%. The prediction effect was good and feasible.

Optimal Medical Image Size Reduction Model Creation Using Recurrent Neural Network and GenPSOWVQ.

Medical diagnosis is always a time and a sensitive approach to proper medical treatment. Automation systems have been developed to improve these issues. In the process of automation, images are processed and sent to the remote brain for processing and decision making. It is noted that the image is written for compaction to reduce processing and computational costs. Images require large storage and transmission resources to perform their operations. A good strategy for pictures compression can help minimize these requirements. The question of compressing data on accuracy is always a challenge. Therefore, to optimize imaging, it is necessary to reduce inconsistencies in medical imaging. So this document introduces a new image compression scheme called the GenPSOWVQ method that uses a recurrent neural network with wavelet VQ. The codebook is built using a combination of fragments and genetic algorithms. The newly developed image compression model attains precise compression while maintaining image accuracy with lower computational costs when encoding clinical images. The proposed method was tested using real-time medical imaging using PSNR, MSE, SSIM, NMSE, SNR, and CR indicators. Experimental results show that the proposed GenPSOWVQ method yields higher PSNR SSIMM values for a given compression ratio than the existing methods. In addition, the proposed GenPSOWVQ method yields lower values of MSE, RMSE, and SNR for a given compression ratio than the existing methods.

IntOPMICM: Intelligent Medical Image Size Reduction Model.

Due to the increasing number of medical imaging images being utilized for the diagnosis and treatment of diseases, lossy or improper image compression has become more prevalent in recent years. The compression ratio and image quality, which are commonly quantified by PSNR values, are used to evaluate the performance of the lossy compression algorithm. This article introduces the IntOPMICM technique, a new image compression scheme that combines GenPSO and VQ. A combination of fragments and genetic algorithms was used to create the codebook. PSNR, MSE, SSIM, NMSE, SNR, and CR indicators were used to test the suggested technique using real-time medical imaging. The suggested IntOPMICM approach produces higher PSNR SSIM values for a given compression ratio than existing methods, according to experimental data. Furthermore, for a given compression ratio, the suggested IntOPMICM approach produces lower MSE, RMSE, and SNR values than existing methods.

Analysis of photostimulable phosphor image plate artifacts and their prevalence.

BACKGROUND Digital radiography has recently been used in dentistry as a substitute for conventional film radiography worldwide. Digital imaging has many advantages and provides new possibilities for recording and interpreting radiographic data. This system uses different types of digital receptors.AIMTo detect the frequency, type, and reasons behind the appearance of intraoral image artifacts acquired by photostimulable phosphor plates (PSP).METHODSThis retrospective descriptive study was conducted in the oral and maxillofacial radiology unit of the dental clinics of the College of Dentistry, Princess Nourah University (PNU). All intraoral digital radiographs were acquired using (Gendex Expert DC., United States) an intraoral X-ray machine with 7 -mA, 65-kVP using a PSP system (Soredex DIGORA Optime imaging plate) and laser scanners (Soredex DIGORA Optime), which can house all sizes of reusable intraoral PSP sensor plates with image acquisition software (MIPACS Dental Enterprise viewer 3.2.2). A total of 50000 intraoral radiographs were retrieved from the clinical database from April 2018 to April 2020 to evaluate the reason, type, and solutions to these image artifacts.RESULTSOverall, 50000 intraoral digital radiographs were acquired in a two-year-period; that is, from April 2018 to April 2020. Of these, 3550 (7.1%) retakes were performed due to the presence of image artifacts. Periapical radiographs were the most common image type of intraoral retakes (80.8%). Imaging artifacts were divided into three categories: operator, plate and scanning errors. Out of 3550 retakes, 5%, 1.37%, and 0.73% were related to operator, plate, and scanning errors, respectively. The cone cut was the most common operator error (988 images), Bite marks were the most common plate error (276 images), and delayed scanning artifacts were the most common scanning errors (145 images). The calculated kappa value for interobserver reliability was 0.99, indicating almost perfect interobserver agreement. CONCLUSIONOur study discussed intraoral image artifacts that were characteristic of PSP, where the most common artifacts were bitemarks, image size reduction, scratches, and delayed scanning.