Kinds Of Images Research Articles

Fighting fire with fire: A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification

Adversarial images generated by generative adversarial networks are not close to any existing benign images, and contain nonrobust features that have been identified as critical to the robustness of a machine learning model. Since adversarial images have an underlying distribution that differs from normal images, these kinds of images can offer valuable features for training a robust model. To deal with these special features, we focus on a novel machine learning task of adversarial images classification, where adversarial images can be used to investigate the problem of classifying adversarial images themselves. In the setting of this novel task, adversarial images are the ONLY kind of data used in training and testing, rather than not just a set of testing images as usual. To this end, we propose a novel spatial–frequency ensemble relation network with generative adversarial learning. First, we present a spatial–frequency ensemble representation learning to extract the feature of training images. Second, we design a meta-learning-based relation model to gain the relationship between images. Third, to achieve a robust model, we utilize generative adversarial learning and transform the relationship into a Jacobian matrix. Finally, we design a discriminator model that determines whether an adversarial image is from the matching category or not. Experimental results demonstrate that our approach achieves significantly higher performance compared with other state-of-the-arts.

Haze concentration adaptive network for image dehazing

Learning-based methods have attracted considerable interest in image dehazing. However, most existing methods are not well adapted to different hazy conditions, especially when dealing with the heavily hazy scene. There is often a significant amount of haze that remains in the images recovered by most methods. To address this issue, we propose an end-to-end Haze Concentration Adaptive Network (HCAN), including a pyramid feature extractor (PFE), a feature enhancement module (FEM), and a multi-scale feature attention module (MSFAM) for image dehazing. Specifically, PFE based on the feature pyramid structure leverages complementary features from different CNN layers to help the clear image prediction. Then, FEM fuses four kinds of images with different haze density (i.e., three recovered images in the FEM with light haze density, and the input hazy image with strong haze condition) to guide the network to adaptively perceive images under different haze conditions. Finally, MSFAM is designed under two principles, multi-scale structure and attention mechanism. It is used to help the network produce a clear image with more details, and ease the network training. Comprehensive experiments demonstrate that the proposed HCAN performs favorably against the state-of-the-art methods in terms of PSNR, SSIM, and visual effect. The results, per-trained models and code are available at https://github.com/TaoWangzj/HCAN.

Intercalibration Between Night-Time DMSP/OLS Radiance Calibrated Images and NPP/VIIRS Images Using Stable Pixels

In this study, we proposed a method for intercalibration between DMSP/OLS radiance calibrated images and NPP/VIIRS images based on stable pixels, while the challenge is that these two datasets do not have overlapped periods which makes it difficult to find the stable pixels. In this study, the outliers, which are the changed pixels, were iteratively removed considering the radiometric and spatial differences between the two kinds of images, and then the stable pixels were obtained. Using the stable pixels, the DMSP/OLS radiance calibrated images and NPP/VIIRS images were intercalibrated based on a regression model. Taking mainland China as an experimental region, we obtained DMSP-like VIIRS images from 2013 to 2019. The growth of night-time light (NTL) from 2010 to later years (e.g., 2013–2019) in provincial regions of mainland China is well correlated to the growth of gross domestic product (GDP), and the averaged ${R^2}$ between the two variables is 0.7126. The growth of NTL from 2010 to later years (e.g., 2013–2019) in 236 municipal regions is also well correlated to the growth of GDP, and the averaged ${R^2} $ is 0.5974. Compared with a multitemporal logistic model which also aims to intercalibrate the two kinds of images, our method showed better a better performance for the intercalibration as our method can generate time series NTL images with stronger correlation to economic growth. Finally, we also found that the NTL, derived from the intercalibrated images, is growing faster in the area where urbanization is faster. This study shows that the proposed method is effective for the intercalibration between the two sets of images which do not have an overlapped time series.

Detecting Adversarial Image Examples in Deep Neural Networks with Adaptive Noise Reduction

Recently, many studies have demonstrated deep neural network (DNN) classifiers can be fooled by the adversarial example, which is crafted via introducing some perturbations into an original sample. Accordingly, some powerful defense techniques were proposed. However, existing defense techniques often require modifying the target model or depend on the prior knowledge of attacks. In this paper, we propose a straightforward method for detecting adversarial image examples, which can be directly deployed into unmodified off-the-shelf DNN models. We consider the perturbation to images as a kind of noise and introduce two classic image processing techniques, scalar quantization and smoothing spatial filter, to reduce its effect. The image entropy is employed as a metric to implement an adaptive noise reduction for different kinds of images. Consequently, the adversarial example can be effectively detected by comparing the classification results of a given sample and its denoised version, without referring to any prior knowledge of attacks. More than 20,000 adversarial examples against some state-of-the-art DNN models are used to evaluate the proposed method, which are crafted with different attack techniques. The experiments show that our detection method can achieve a high overall F1 score of 96.39% and certainly raises the bar for defense-aware attacks.

Journal of the Society for Information Display

Jiun-Haw Lee Editor in Chief The November 2020 issue of the Journal of the Society for InformationDisplay(JSID) contains five regular papers and one letter. In the December issue, there are two invited papers, four regular papers,and two letters. Along with hardware research in devices and systems of displays, the issue also publishes several papers about software progress in displays and visions in algorithm and human-machine interaction (HMI). It also relates to the study in psychophysics, such as visual comfort, and perception. We are excited to see the broadened spectrum of display-related research published in JSID. Amid the COVID-19 pandemic, in-person communication has diminished. To compensate, there has been a substantial uptick in online meetings that require additional displays and better user experiences. As our field responds to the increased demands on display technologies, we welcome any advances in display-related research that elevates human interactions to a better new “normal.” We hope innovators will submit such advancements to JSID. Two invited papers were published in the December 2020 issue of JSID. In the first paper, the authors thoroughly discussed a display system's optical films for application in vehicles. This technology must consider all the safety and convenience issues that occur under harsh environments, such as high temperatures and a wide range of ambient lighting conditions. In the second invited paper, an 8-inch transparent and flexible display was demonstrated with miniLED technology and an indium-gallium-zinc-oxide, thin-film transistor (IGZO-TFT) backplane on the flexible substrate. JSID is published monthly by Wiley. Subscription fees apply, but SID members and student members have free online access via sid.org/Publications.aspx. Many universities also have an institutional subscription. JSID is indexed in the Web of Science. Submit your original manuscript online via mc.manuscriptcentral.com/sid. Author guidelines can be found on the Journal's homepage: tinyurl.com/jsidhome. Editorial board: tinyurl.com/jsideb. Please direct any questions about JSID to its editor in chief at editor@sid.org. EarlyView: accepted papers about to be published can be accessed online via tinyurl.com/jsidev. Image quality equivalence between peak luminance and chromaticity gamut | Yongmin Park and Michael J. Murdoch | doi: 10.1002/jsid.906 In this article, psychophysical research was conducted to demonstrate the tradeoff between the image quality's lightness (L) and chroma (C). Essentially, peak luminance should decrease to obtain the same image quality when the color gamut increases. In addition, the regions of interest (ROI) for various kinds of images also differ, which leads to slight alterations in the tradeoff relations. Although the experiment was conducted in a standard dynamic range (SDR) display, this line of research may be extend the concept to a high dynamic range (SDR) display in the future. Highly transparent, ultra-thin flexible, full-color mini-LED display with indium–gallium–zinc oxide thin-film transistor substrate| Yang Sun et al.| doi: 10.1002/jsid.970 As the title of this paper indicates, the authors demonstrated a full-color, transparent, and flexible display by integrating several technologies, including miniLED as the emitter, IGZO as the TFT backbone, and colorless polyimide (CPI) as the transparent and flexible substrate. The display size is 8 inches with a resolution of 46 pixels per inch (ppi) and thickness of 0.2 millimeters (mm). Optical transmittance is 60 percent. Full-width-at half-maximum (FWHM) of LED emission spectra vary from 12 to 27 nanometers (nm), resulting the color gamut of 80 percent in the color-range standard BT 2020. It successfully proves the process integration for miniLED and microLED, which also can be applied to large-size displays. The Journal is soliciting review papers on any display-related topic. If you have a great idea for a review paper, please contact the editor at editor@sid.org. Page charges for invited review papers will be waived.

Image plagiarism detection using GAN - (Generative Adversarial Network)

In Today’s date plagiarism is a very important aspect because content originality is the client's prior requirement. Many people on the internet use others' images and get publicity while the owner of the image or data won′t get anything out of it. Many users copy the data or image features from the other users and modify it a little bit or create an artificial replica of it. With sufficient computational power and volume of data, the GAN models are capable enough to produce fake images that look very much similar to the real images. These kinds of images are generally not detected by modern plagiarism systems. GAN stands for generative adversarial network. It has two neural networks working inside. The first one is the generator which generates a random image and the second one is the discriminator which identifies whether the image being generated is a real or a fake image. In this paper, we have proposed a system that has been trained on both fake images (GAN Generated images) and real images and will help us in flagging whether the image is plagiarised or a real image.

A Comparative Study of Stand-Alone and Hybrid CNN Models for COVID-19 Detection

The COVID-19 pandemic continues to impact both the international economy and individual lives. A fast and accurate diagnosis of COVID-19 is required to limit the spread of this disease and reduce the number of infections and deaths. However, a time consuming biological test, Real-Time Reverse Transcription–Polymerase Chain Reaction (RT-PCR), is used to diagnose COVID-19. Furthermore, sometimes the test produces ambiguous results, especially when samples are taken in the early stages of the disease. As a potential solution, machine learning algorithms could help enhance the process of detecting COVID-19 cases. In this paper, we have provided a study that compares the stand-alone CNN model and hybrid machine learning models in their ability to detect COVID-19 from chest X-Ray images. We presented four models to classify such kinds of images into COVID-19 and normal. Visual Geometry Group (VGG-16) is the architecture used to develop the stand-alone CNN model. This hybrid model consists of two parts: the VGG-16 as a features extractor, and a conventional machine learning algorithm, such as support-vector-machines (SVM), Random-Forests (RF), and Extreme-Gradient-Boosting (XGBoost), as a classifier. Even though several studies have investigated this topic, the dataset used in this study is considered one of the largest because we have combined five existing datasets. The results illustrate that there is no noticeable improvement in the performance when hybrid models are used as an alternative to the stand-alone CNN model. VGG-16 and (VGG16+SVM) models provide the best performance with a 99.82% model accuracy and 100% model sensitivity. In general, all the four presented models are reliable, and the lowest accuracy obtained among them is 98.73%.

Zarządzanie w antropocenie, czyli co w czym i co z tego. Analiza literatury dotyczącej splotu zjawisk

Management in the Anthropocene and What Comes Out of It. Analysis of the Literature on the Entanglement of Phenomena The Anthropocene is a term that not only conjures up all kinds of images in the mind but provides an impulse to reconsider the scope of human responsibility for man-triggered processes and its place in the system of related factors on our planet. At the same time, it is a term treated by many with ambivalence, reluctance, and caution, as it often harbingers the imminent environmental doom, the awareness of which may change the current balance of political and economic forces. Additionally, it is still involved in emphasizing the central role of the human, it is sometimes romanticized and can lead to an aestheticization of the climate catastrophe instead of taking actions resulting not from the will of heroism, but from humility. The ongoing debate on the Anthropocene in the field of management studies is of extraordinary importance, because it provides a framework for undertaking any activity – the activity which either aggravates or alleviates the negative environmental impact. The purpose of this paper is to discuss the contexts in which the scientific literature in the field of management talks about the Anthropocene and to explore the level of gravity that is assigned to management in this conventional geological and cultural era. Particular attention is paid to the dominant trends of reflection which illustrate a wide variety of attitudes towards the Anthropocene, including the one that places the Anthropocene against the background of efforts to maintain the status quo and the one that perceives it as a prelude to concocting alternative or even anarchist visions of management. The paper focuses on theoretical voices, which determined the method of analysis based on the study of language and the interpretation of narratives and metaphors.

Playful Letters: A Study in Early Modern Alphabetics by Erika Mary Boeckeler

Reviewed by: Playful Letters: A Study in Early Modern Alphabetics by Erika Mary Boeckeler Adam Hembree Boeckeler, Erika Mary, Playful Letters: A Study in Early Modern Alphabetics, Iowa City, University of Iowa Press, 2017; paperback; pp. 308; 8 colour plates; R.R.P. US $75.00; ISBN 9781609384746. Playful Letters offers a compelling reading of early modern alphabetics, navigating a diverse set of primers, prints, and images spanning from western Europe to Russia. Boeckeler's driving premise—that letters are material bodies with pluripotent histories—provides a persuasive framework for understanding how teachers, artists, and religious or political authorities could wield them. The first chapter engages with French engraver Geofroy Tory, whose seminal work Champ Fleury (1529) anatomizes each letter as a body on a grid, reading each according to didactic humanist principles. Scholars interested in Jeffrey Masten's work on this subject in Queer Philology (University of Pennsylvania Press, 2016) can find a complementary but divergent reading on Fleury's naughty letter 'Q', which Boeckeler takes as a rule rather than an exception to the logic of the cross row. She contends that Q's oddness reflects similar idiosyncrasies across the Fleury alphabet, and that many letters have the potential for disrupting his apparent project of language standardization. This is precisely what renders these letters as 'playful' in Boeckeler's view. From Fleury in France, we move to Peter Flötner's Menschenalphabet in Germany and I. Paulini's Ovidian Alphabet in Italy. In these and comparable examples, Boeckeler encourages readers to perform complementary but inverted practices: read printed text as a kind of image, and read images (for example, Flötner's human bodies posed as letters) as text. Chapter 3 follows this train of thought in using signatures and typography to interpret Albrecht Dürer's self-portrait by superimposition. The relationships between letters on Flötner's page, as well as the visual echoes across these alphabets and prints, are all extremely intriguing. In these two chapters more than most, however, Boeckeler shows a habit of stretching explanatory metaphors beyond their tension limits, which can at times obscure opportunities for clear conclusions. Shakespeare's Richard III presents a prophesy in which a certain 'G' that is plucked from the cross row will disinherit a king. In her fourth chapter, Boeckeler closely reads the possible valences of this moment on stage, with varying levels of success. Her link between Richard's 'G' and the central 'G' of a cruciform cross row is an exciting one, while the assertions that either the two Gs in the written form of 'George' or Richard's 'G-shaped' body would affect readers or audiences is tenuous. The chapter resolves with the much more compelling [End Page 197] claim that Titus Andronicus centres the Roman alphabet as a site and preserver of imperial violence. Boeckeler's earlier reference to 'onomancy' (the magic of names) contributes to this to build a conclusion that speech act theory has long used as a founding premise: speech is an action. The final chapter of Playful Letters offers a fascinating historical connection between this ostensibly western European history of alphabetics and complementary developments in Russia over the same period. Scholars of early modern letters as pedagogy in England, for example, will no doubt find the history of the premodern Cyrillic letter Az extremely compelling, with its simultaneous roles as first letter and first-person pronoun. Boeckeler guides the reader through the longstanding theological and philosophical resonances possible in an alphabet—az-buka—that literally means 'I am books' or, more pertinently to Boeckeler's premise: 'I am letters'. The Cyrillic alphabet's evolution and historic idiosyncrasies offer a divergent, but familiar, promise of the letter as a lens through which to make, see, and understand the self. Playful Letters offers an excellent point of reference for a range of both touchstone texts and lesser-known works on the early modern Roman and Cyrillic alphabets. It offers a wealth of rigorous research and intriguing intertextual connections to inspire further inquiry on the use of letters. Some of its theoretical framing work—especially the notion of 'letterature' as a term for period discourses on letters—ultimately proves an unnecessary...

The impact of interdisciplinary dynamic images on public perception

Dynamic image, usually defined as motion graphics or dynamic graphics, is a kind of image art based on art design and computer science. In a broad sense, dynamic image is a discipline that integrates animation, film, and graphic design. Its expression content is more refined and straightforward than animated movies. It is more comprehensive than graphic design in conveying information. There are more types of dynamic images in the new media era. There are virtual reality technologies that rely on computer science and interactivity, and there are self-media on the Internet. Together with traditional media, they have more and more impact on the public's cognition. The interdisciplinary dynamic image is a visual language based on the dynamic image and interdisciplinary boundaries. It is main feature is to use dynamic images to express the main content of something, event or thought, a moving image made for the purpose of improving the public’s understanding of the corresponding thing, In the interdisciplinary dynamic images, the concreteness and narrative nature of the dynamic images are weakened, and the non-representation and scientific are more emphasized.

Technologies of Representation

“In teaching us a new visual code”, Susan Sontag wrote more than forty years ago, “photographs alter and enlarge our notions of what is worth looking at and what we have a right to observe” (Sontag, 1977, p. 3). Since the publication of Sontag’s `Plato’s Cave ?, the most radical change in this visual code has been the pace and breadth of its reach. We carry image-making and image-sharing devices in a pocket. We can easily upload a photograph to a search engine and call up millions of images based on visual or conceptual similarity. But even with the leveling of tools for making and distributing images, Sontag’s empowering `ethics of seeing ? is today a territory in dispute. Millions of image-makers and instant sharing capabilities are met by algorithmic filter-bubbles and widespread misinformation campaigns. The sheer quantity of image circulation did not amount to improved visibility, let alone mutual understanding.
 After 18-O (18 October) in Chile and the Covid-19 pandemic, it became clear that any discussion on representation should address the unrealized promise of the `ethics of images ?. This implies not forgetting that, as Gayatri Spivak proposes, whenever we use the word ‘representation’ we are compounding its two meanings: to ‘re-present’ as in art or philosophy, and to ‘speak for’ as in politics (1988, p. 275). Architects are familiar with images that ‘do’ things for us: renders pre-visualize, orthographic projections measure, collages hint at experiences of space. Architectural drawings can be translated, as Robin Evans (1997) put it, into buildings and urban plans. But what is architecture’s relationship to other kinds of images, those which were never meant to become buildings? This issue of Materia Arquitectura was a call to explore the agency of images in the construction of realities, the imposition of borders and the narration of stories that are political. Not only because their object is the polis, but because they alter our relationship with the built environment and consequently, the way in which we understand, imagine and shape, as a society, the common territory that is at the base of the exercise of public power.

A Review on LBP in Image Retrieval System for Future Enhancement and Vector Images

Content-Based Image Retrieval is a process to retrieve the similar images from the large set of image database corresponding to the query image. In CBIR low level or pixel level features such as color, texture and shape of the images are extracted and on the basis of similarity matching algorithm the required similar kind of images are retrieved from the image database. To understand the evaluation and evolution of CBIR system various research was studied and various research is going on this way also. In this paper, we have discussed some of the popular pixel level feature extraction techniques for Content-Based Image Retrieval and we also present here about the performance of each technique.

Science and the Pragmatist Image of Humanity: Lessons from Wilfrid Sellars and Beyond

Abstract The paper focuses on the pragmatist image of humanity based on a re-reading of the philosophical “manifesto” of Wilfrid Sellars (1963) in which he became entangled in the dichotomy between “scientific” and “manifest” images. The key to solving this problem, according to the author, is the new pragmatist understanding of science as a cultural practice, which provide us with a new framework for transcending this dichotomy. By reconstructing Sellars in an anthropological rather than a scientistic way and by drawing on humanistic philosophical intentions that are present both in pragmatism and in Sellars, it becomes possible to outline a concept of “science with a human face.” The purpose of all kinds of images, including scientific ones, is to serve the enrichment of human understanding and life.

Personality Trait Analysis in Social Networks Based on Weakly Supervised Learning of Shared Images

Social networks have attracted the attention of psychologists, as the behavior of users can be used to assess personality traits, and to detect sentiments and critical mental situations such as depression or suicidal tendencies. Recently, the increasing amount of image uploads to social networks has shifted the focus from text to image-based personality assessment. However, obtaining the ground-truth requires giving personality questionnaires to the users, making the process very costly and slow, and hindering research on large populations. In this paper, we demonstrate that it is possible to predict which images are most associated with each personality trait of the OCEAN personality model, without requiring ground-truth personality labels. Namely, we present a weakly supervised framework which shows that the personality scores obtained using specific images textually associated with particular personality traits are highly correlated with scores obtained using standard text-based personality questionnaires. We trained an OCEAN trait model based on Convolutional Neural Networks (CNNs), learned from 120K pictures posted with specific textual hashtags, to infer whether the personality scores from the images uploaded by users are consistent with those scores obtained from text. In order to validate our claims, we performed a personality test on a heterogeneous group of 280 human subjects, showing that our model successfully predicts which kind of image will match a person with a given level of a trait. Looking at the results, we obtained evidence that personality is not only correlated with text, but with image content too. Interestingly, different visual patterns emerged from those images most liked by persons with a particular personality trait: for instance, pictures most associated with high conscientiousness usually contained healthy food, while low conscientiousness pictures contained injuries, guns, and alcohol. These findings could pave the way to complement text-based personality questionnaires with image-based questions.

Soft Compression: An Approach to Shape Coding for Images

In this letter, we propose soft compression, an lossless compression approach to shape coding for images using location index and codebook of designed shapes with various sizes. This method is different from traditional image compression methods, as it aims at finding the basic shape blocks of pictures to improve the compression ratio from the perspective of information theory and frequency. In particular, the triplets including location and codeword are stored when encoding an image instead of the entire array. It is shown that soft compression can greatly reduce the bandwidth and storage space needed in the process of transmitting and storing the same kind of images.

Entropy estimation for robust image segmentation in presence of non Gaussian noise

In this work we introduce a new approach for robust image segmentation. The idea is to combine two strategies within a Bayesian framework. The first one is to use a Markov Random Field (MRF), which allows to introduce prior information with the purpose of preserve the edges in the image. The second strategy comes from the fact that the probability density function (pdf) of the likelihood function is non Gaussian or unknown, so it should be approximated by an estimated version, and for this, it is used the classical non-parametric or kernel density estimation. This two strategies together lead us to the definition of a new maximum a posteriori (MAP) approach based on the minimization of the entropy of the estimated pdf of the likelihood function and the MRF at the same time, named MAP entropy estimator (MAPEE). Some experiments were conducted for different kind of images degraded with impulsive noise and other non-Gaussian distributions, where the segmentation results are very satisfactory comparing them with respect to recent robust approaches based on the fuzzy c-means (FCM) segmentation.

Religious façade of ‘the Chinese nation’ in China’s school curriculum

ABSTRACT The shaping of people’s identities as members of the Chinese nation is crucial to China’s nationalist discourse at school. What kind of image of the Chinese nation does the state intend to transmit to its new generation of citizens? Does the constitutional separation of religion and education prescribe a secular image for the collective identity of the Chinese nation? This study explores how the interpretation of religion is manufactured during the discourse of constructing the nation in Chinese school textbooks. The findings reveal that different religious traditions are assigned strikingly different roles in the national curriculum’s interpretation of the Chinese nation as a construct and the distinction between China and the external world. The integration of religious narratives and the agenda of shaping the national cultural identity of Chinese students reveal the changing role of religion in contemporary Chinese society.

‘Behind the Digi-God's back’: social representations of older people's digital competences and internet use in regional Finnish newspapers

AbstractThis article discusses how three Finnish regional newspapers represented older people's digital competences and internet use in their daily coverage. The study explored media representations from the perspective of social representations and sought to answer the following questions: In what kind of internet user roles do the articles portray older people? How and with what kind of images do the articles portray older people's digital competences in various internet user roles? How are older people positioned at a societal level in the articles? The analysis revealed that older people were portrayed as incompetent outcasts of a digitalised society. However, there was a clear difference according to whether older people were portrayed as recipients of public services or as consumers of private services. As targets of public services, older people were predominantly portrayed as happy targets, who welcomed the services provided for them. This result can be interpreted as part of the promotion of government digitalisation policies.

Quality and texture analysis of biometric images compressed with second-generation wavelet transforms and SPIHT-Z encoder

<span>In biometric systems, compression takes important place especially in order to reduce the size of the information stored or transmitted through the distributed biometric systems. It is also noted that the compression techniques induce loss of information in the compressed images that can affect the effectiveness of biometric systems. The main objective of our contribution is to examine the efficacy of the used method to offer an optimal compression quality in these kind of images without considerable distortion. In order to evaluate the efficacy of the compression process, we use two kinds of evaluation, full-reference image quality assessment and a new proposed textural quality analysis of the compressed images. In this paper, we use a second-generation wavelet transform to improve the compression study in biometric images. The basic idea of this algorithm is the quincunx wavelet transform coupled to a modified progressive encoder called SPIHT-Z encoding.</span>

High-Precision Spotlight SAR Imaging with Joint Modified Fast Factorized Backprojection and Autofocus

Fast factorized backprojection (FFBP) takes advantage of high accuracy of time-domain algorithms while also possessing high efficiency comparable with conventional frequency domain algorithms. When phase errors need to be compensated for high-resolution synthetic aperture radar (SAR) imaging, however, neither polar formatted subimages within FFBP flow nor the final Cartesian image formed by FFBP is suitable for phase gradient autofocus (PGA). This is because these kinds of images are not capable of providing PGA with a clear Fourier transform relationship (FTR) between image domain and range-compressed phase history domain. In this paper, we make some essential modifications to the original FFBP and present a scheme to incorporate overlapped-subaperture frame for an accurate PGA processing. The raw data collected by an airborne high-resolution spotlight SAR are used to demonstrate the performance of this algorithm.