Showing papers on "Aerial image published in 2009"

Change Detection in Optical Aerial Images by a Multilayer Conditional Mixed Markov Model

Abstract: In this paper, we propose a probabilistic model for detecting relevant changes in registered aerial image pairs taken with the time differences of several years and in different seasonal conditions. The introduced approach, called the conditional mixed Markov model, is a combination of a mixed Markov model and a conditionally independent random field of signals. The model integrates global intensity statistics with local correlation and contrast features. A global energy optimization process ensures simultaneously optimal local feature selection and smooth observation-consistent segmentation. Validation is given on real aerial image sets provided by the Hungarian Institute of Geodesy, Cartography and Remote Sensing and Google Earth.

Vision-based unmanned aerial vehicle navigation using geo-referenced information

Abstract: This paper investigates the possibility of augmenting an Unmanned Aerial Vehicle (UAV) navigation system with a passive video camera in order to cope with long-term GPS outages. The paper proposes a vision-based navigation architecture which combines inertial sensors, visual odometry, and registration of the on-board video to a geo-referenced aerial image. The vision-aided navigation system developed is capable of providing high-rate and drift-free state estimation for UAV autonomous navigation without the GPS system. Due to the use of image-to-map registration for absolute position calculation, drift-free position performance depends on the structural characteristics of the terrain. Experimental evaluation of the approach based on offline flight data is provided. In addition the architecture proposed has been implemented on-board an experimental UAV helicopter platform and tested during vision-based autonomous flights.

Generation of aerial images

Abstract: The method according to the invention gene rates an aerial image mosaic viewing a larger area than a single image from a camera can provide using a combination of computer vision and photogrammetry. The aerial image mosaic is based on a set of images acquired from a camera. Selective matching and cross matching of consecutive and non-consecutive images, respectively, are performed and three dimensional motion and structure parameters are calculated and implemented on the model to check if the model is stable. Thereafter th e parameters are globally optimised and based on these optimised parameters the aerial image mosaic is generated. The set of images may be limited by removing old image data as new images are acquired. The method makes it is possible to establish images in near real time using a system of low complexity and small size, and using only image information.

Automatic Vegetation Identification and Building Detection from a Single Nadir Aerial Image

Abstract: A novel, automatic tertiary classifier is proposed for identifying vegetation, building and non-building objects from a single nadir aerial image. The method is unsupervised, that is, no parameter adjustment is done during the algorithm’s execution. The only assumption the algorithm makes about the building structures is that they have convex rooftop sections. Results are provided for two different actual data sets.

Semantic classification in aerial imagery by integrating appearance and height information

Abstract: In this paper we present an efficient technique to obtain accurate semantic classification on the pixel level capable of integrating various modalities, such as color, edge responses, and height information. We propose a novel feature representation based on Sigma Points computations that enables a simple application of powerful covariance descriptors to a multi-class randomized forest framework. Additionally, we include semantic contextual knowledge using a conditional random field formulation. In order to achieve a fair comparison to state-of-the-art methods our approach is first evaluated on the MSRC image collection and is then demonstrated on three challenging aerial image datasets Dallas, Graz, and San Francisco. We obtain a full semantic classification on single aerial images within two minutes. Moreover, the computation time on large scale imagery including hundreds of images is investigated.

Individual Tree Species Classification by Illuminated—Shaded Area Separation

Abstract: A new method, called Illumination Dependent Colour Channels (IDCC), is presented to improve individual tree species classification. The method is based on tree crown division into illuminated and shaded parts on a digital aerial image. Colour values of both sides of the tree crown are then used in species classification. Tree crown division is achieved by comparing the projected location of an aerial image pixel with its neighbours on a Canopy Height Model (CHM), which is calculated from a synchronized LIDAR point cloud. The sun position together with the mapping aircraft position are also utilised in illumination status detection. The new method was tested on a dataset of 295 trees and the classification results were compared with ones measured with two other feature extraction methods. The results of the developed method gave a clear improvement in overall tree species classification accuracy.

Improving Measurement of Forest Structural Parameters by Co-Registering of High Resolution Aerial Imagery and Low Density LiDAR Data.

Abstract: Forest structural parameters, such as tree height and crown width, are indispensable for evaluating forest biomass or forest volume. LiDAR is a revolutionary technology for measurement of forest structural parameters, however, the accuracy of crown width extraction is not satisfactory when using a low density LiDAR, especially in high canopy cover forest. We used high resolution aerial imagery with a low density LiDAR system to overcome this shortcoming. A morphological filtering was used to generate a DEM (Digital Elevation Model) and a CHM (Canopy Height Model) from LiDAR data. The LiDAR camera image is matched to the aerial image with an automated keypoints search algorithm. As a result, a high registration accuracy of 0.5 pixels was obtained. A local maximum filter, watershed segmentation, and object-oriented image segmentation are used to obtain tree height and crown width. Results indicate that the camera data collected by the integrated LiDAR system plays an important role in registration with aerial imagery. The synthesis with aerial imagery increases the accuracy of forest structural parameter extraction when compared to only using the low density LiDAR data.

Drawing stereo disparity images into occupancy grids: Measurement model and fast implementation

Abstract: Mapping the environment is necessary for navigation in unknown areas with autonomous vehicles. In this context, a method to process depth images for occupancy grid mapping is developed. Input data are images with pixel-based distance information and the corresponding camera poses. A measurement model, focusing on stereo-based depth images and their characteristics, is presented. Since an enormous amount of range data must be processed, improvements like image pyramids are used so that the image analysis is possible in real-time. Output is a grid-based image interpretation for sensor fusion, i.e. a world-centric occupancy probability array containing information stored in a single image. Different approaches to draw pixel information into a grid map are presented and discussed in terms of accuracy and performance. As a final result, 3D occupancy grids from aerial image sequences are presented.

Large scale graph-based SLAM using aerial images as prior information

Abstract: To effectively navigate in their environments and accurately reach their target locations, mobile robots require a globally consistent map of the environment. The problem of learning a map with a mobile robot has been intensively studied in the past and is usually referred to as the simultaneous localization and mapping (SLAM) problem. However, existing solutions to the SLAM problem typically rely on loop-closures to obtain global consistency and do not exploit prior information even if it is available. In this paper, we present a novel SLAM approach that achieves global consistency by utilizing publicly accessible aerial photographs as prior information. Our approach inserts correspondences found between three-dimensional laser range scans and the aerial image as constraints into a graph-based formulation of the SLAM problem. We evaluate our algorithm based on large real-world datasets acquired in a mixed in- and outdoor environment by comparing the global accuracy with state-of-the-art SLAM approaches and GPS. The experimental results demonstrate that the maps acquired with our method show increased global consistency.

Detection of Object Motion Regions in Aerial Image Pairs With a Multilayer Markovian Model

Abstract: We propose a new Bayesian method for detecting the regions of object displacements in aerial image pairs. We use a robust but coarse 2D image registration algorithm. Our main challenge is to eliminate the registration errors from the extracted change map. We introduce a three-layer Markov random field (L 3MRF) model which integrates information from two different features, and ensures connected homogenous regions in the segmented images. Validation is given on real aerial photos.

Tree detection from aerial imagery

Abstract: We propose an automatic approach to tree detection from aerial imagery. First a pixel-level classifier is trained to assign a {tree, non-tree} label to each pixel in an aerial image. The pixel-level classification is then refined by a partitioning algorithm to a clean image segmentation of tree and non-tree regions. Based on the refined segmentation results, we adopt template matching followed by greedy selection to locate individual tree crowns.As training a pixel-level classifier requires manual generation of ground-truth tree masks, we propose methods for automatic model and training data selection to minimize the manual work and scale the algorithm to the entire globe. We test the algorithm on thousands of production aerial images across different countries. We demonstrate high-quality tree detection results as well as good scalability of the proposed approach.

Evaluation of the self‐organizing map classifier for building detection from lidar data and multispectral aerial images

Abstract: Integration of aerial images and lidar data compensate for the individual weaknesses of each data set when used alone, thus providing more accurate classification of terrain cover, such as buildings, roads and green areas, and advancing the potential for automation of large scale digital mapping and GIS database compilation. This paper presents work on the development of automatic feature extraction from multispectral aerial images and lidar data. A total of 22 feature attributes have been generated from the aerial image and the lidar data which contribute to the detection of the features. The attributes include those derived from the Grey Level Cooccurrence Matrix (GLCM), Normalized Difference Vegetation Indices (NDVI), and standard deviation of elevations and slope. A Self‐Organizing Map (SOM) was used for fusing the aerial image, lidar data and the generated attributes for building detection. The classified images were then processed through a series of image processing techniques to separate the detec...

Mask defect auto disposition based on aerial image in mask product

Abstract: At the most advanced technology nodes, such as 45nm and below, aggressive OPC and Sub-Resolution Assist Features (SRAFs) are required. However, their use results in significantly increased mask complexity, making mask defect disposition more challenging than ever. In an attempt to mitigate such difficulties, new mask inspection technologies that rely on hardware emulation and software simulation to obtain aerial image at the wafer plane have been developed; however, automatic mask disposition based on aerial image is still problematic because aerial image does not give the final resist CD or contour, which are commonly used in lithography verification on post OPC masks. In this paper, an automated mask defect disposition system that remedies these shortcomings is described. The system, currently in use for mask production, works in both die-to-die and die-to-database modes, and can operate on aerial images from both AIMSTM and aerial-image-based inline mask inspection tools. The disposition criteria are primarily based on waferplane CD variance. The system also connects to a post-OPC lithography verification tool that can provide gauges and CD specs, which are then used in the mask defect disposition.

Aerial image generating apparatus, aerial image generating method, and storage medium having aerial image generating program stored therein

Abstract: An apparatus and method generating a road image including no features such as trees and tunnels hiding or covering a road surface. A mobile measuring apparatus installed in a vehicle may acquire a distance and orientation point cloud, a camera image, GPS observation information, a gyro measurement value, and an odometer measurement value, while moving in a target area. The position and attitude localizing apparatus may localize the position and attitude of the vehicle based on the GPS observation information, the gyro measurement value and the odometer measurement value. The point cloud generating apparatus may generate a point cloud based on the camera image, the distance and orientation point cloud, and a position and attitude localized value. The point cloud orthoimage generating apparatus may extract points close to a road surface exclusively from the point cloud by removing points higher than the road surface, orthographically project each extracted point onto a horizontal plane, and generate a point cloud orthoimage. The point cloud orthoimage may show the road surface including no features covering or hiding the road surface.

Ultimate contact hole resolution using immersion lithography with line/space imaging

Abstract: Contact Hole (CH) resolution is limited by the low aerial image contrast using dark field masks. Moreover the 2- Dimensional character of CH is a limiting factor in the use of extreme Resolution Enhancement Techniques for reaching the smallest pitch. These limitations can be overcome if one deconvolves the 2D CH into two exposures of 1D structures (i.e. lines). These 1D structures can indeed be printed at the ultimate resolution limit of the scanner using dipole exposures. Recently, several materials have become available to pattern CH from such a double exposure of line patterns. It is shown in this paper how this concept of deconvolution can be used in different techniques: Two 1D aerial images can be recomposed in order to obtain 2D images which will subsequently be reversed into CH. We can distinguish, on the one hand, a reversal based on the positive development of line crossings into resist pillar patterns, on which are deposited or coated a gap-fill material layer. The pillars are then removed, leaving a masking material layer with holes. On the other hand, negative tone development can be used to reverse directly the recomposed 2D aerial image: while the classical positive development creates pillars, the negative tone development inverses immediately this image to create contact holes in the resist layer. In this paper, we demonstrate the potential of the double exposure method. We characterise three reversal techniques using a NA=1.35 immersion scanner for patterning 40nm or lower CH at pitch 80nm. We also show etch performance of these processes and address the complexity of each solution.

Fast Vehicle Detection and Tracking in Aerial Image Bursts

Abstract: Caused by the rising interest in traffic surveillance for simulations and decision management many publications concentrate on automatic vehicle detection or tracking. Quantities and velocities of different car classes form the data basis for almost every traffic model. Especially during mass events or disasters a wide-area traffic monitoring on demand is needed which can only be provided by airborne systems. This means a massive amount of image information to be handled. In this paper we present a combination of vehicle detection and tracking which is adapted to the special restrictions given on image size and flow but nevertheless yields reliable information about the traffic situation. Combining a set of modified edge filters it is possible to detect cars of different sizes and orientations with minimum computing effort, if some a priori information about the street network is used. The found vehicles are tracked between two consecutive images by an algorithm using Singular Value Decomposition. Concerning their distance and correlation the features are assigned pairwise with respect to their global positioning among each other. Choosing only the best correlating assignments it is possible to compute reliable values for the average velocities.

Aerial image based technique for measurement of lens aberrations up to 37th Zernike coefficient in lithographic tools under partial coherent illumination.

Abstract: This paper proposes a technique for in situ measurement of lens aberrations up to the 37th Zernike coefficient in lithographic tools under partial coherent illumination. The technique requires the acquisition and analysis of aerial image intensities of a set of 36 binary gratings with different pitches and orientations. By simplifying the theoretical derivation of the optical imaging under partial coherent illumination, two linear models are proposed in a compact expression with two matrixes, which can be easily obtained in advance by numerical calculation instead of by lithographic simulators, and then used to determine the Zernike coefficients of odd aberration and even aberration respectively. The simulation work conducted by PROLITH has validated the theoretical derivation and confirms that such a technique yields a superior quality of wavefront estimate with an accuracy of Zernike coefficients on the order of 0.1 m lambdas (lambda = 193 nm) and an accuracy of wavefronts on the order of m lambdas, due to further considering the influence of the partial coherence factor on pupil sampling. It is fully expected that this technique will simple to implement and will provide a useful practical means for the in-line monitoring of imaging quality of lithographic tools under partial coherent illumination.

Improving the performance of the actinic inspection tool with an optimized alignment procedure

Abstract: Extreme ultraviolet (EUV) microscopy is an important tool for the investigation of the performance of EUV masks, for detecting the presence and the characteristics of defects, and for evaluating the effectiveness of defect repair techniques. Aerial image measurement bypasses the difficulties inherent to photoresist imaging and enables high data collection speed and flexibility. It provides reliable and quick feedback for the development of masks and lithography system modeling methods. We operate the SEMATECH Berkeley Actinic Inspection Tool (AIT), a EUV microscope installed at the Advanced Light Source at Lawrence Berkeley National Laboratory. The AIT is equipped with several high-magnification Fresnel zoneplate lenses, with various numerical aperture values, that enable it image the reflective mask surface with various resolution and magnification settings. Although the AIT has undergone significant recent improvements in terms of imaging resolution and illumination uniformity, there is still room for improvement. In the AIT, an off-axis zoneplate lens collects the light coming from the sample and an image of the sample is projected onto an EUV-sensitive CCD camera. The simplicity of the optical system is particularly helpful considering that the AIT alignment has to be performed every time that a sample or a zoneplate is replaced. The alignment is sensitive to several parameters such as the lens position and orientation, the illumination direction and the sample characteristics. Since the AIT works in high vacuum, there is no direct access to the optics or to the sample during the alignment and the measurements. For all these reasons the alignment procedures and feedback can be complex, and in some cases can reduce the overall data throughput of the system. In this paper we review the main strategies and procedures that have been developed for quick and reliable alignments, and we describe the performance improvements we have achieved, in terms of aberration magnitude reduction.

Normalized Image Log Slope with Secondary Electron Migration Effect in Chemically Amplified Extreme Ultraviolet Resists

Abstract: In lithography, the normalized image log slope (NILS) is an important metric that describes the quality of an aerial image of incident photons. The chemical gradient is also an important metric that describes the quality of a latent image in terms of line edge roughness. The relationship between NILS and the chemical gradient has been theorized in studies on photolithography. In extreme ultraviolet (EUV) resists, however, secondary electrons contribute to the image formation in contrast to the case of photoresists. In this study, we proposed a NILS in which secondary electron migration is taken into account.

Utilizing prior information to enhance self-supervised aerial image analysis for extracting parking lot structures

Abstract: Road network information (RNI) simplifies autonomous driving by providing strong priors about driving environments. Its usefulness has been demonstrated in the DARPA Urban Challenge. However, the need to manually generate RNI prevents us from fully exploiting its benefits. We envision an aerial image analysis system that automatically generates RNI for a route between two urban locations. As a step toward this goal, we present an algorithm that extracts the structure of a parking lot visible in an aerial image. We formulate this task as a problem of parking spot detection because extracting parking lot structures is closely related to detecting all of the parking spots. To minimize human intervention in use of aerial imagery, we devise a self-supervised learning algorithm that automatically obtains a set of canonical parking spot templates to learn the appearance of a parking lot and estimates the structure of the parking lot from the learned model. The data set extracted from a single image alone is too small to sufficiently learn an accurate parking spot model. To remedy this insufficient positive data problem, we utilize self-supervised parking spots obtained from other aerial images as prior information and a regularization technique to avoid an overfitting solution.

Automated electric utility pole detection from aerial images

Abstract: This paper presents an algorithm for the recognition of similar electrical poles from an aerial image by detecting the pole shadow. One pole is used as a template (already identified by a human operator) for the algorithm. The algorithm includes feature extraction, candidate position determination, and elimination of redundant candidates. First, features of a pole shadow are extracted using standard filters and image processing techniques. Then the extracted features are used to design convolution filters tailored to emphasize possible locations for the shadows. Subsequently, an image candidate is submitted to Radon Transformation to verify adherence to expected shadow characteristics. Simulations show that most poles are made much more noticeable by the algorithm.

Self-supervised aerial image analysis for extracting parking lot structure

Abstract: Road network information simplifies autonomous driving by providing strong priors about environments. It informs a robotic vehicle with where it can drive, models of what can be expected, and contextual cues that influence driving behaviors. Currently, however, road network information is manually generated using a combination of GPS survey and aerial imagery. These manual techniques are labor intensive and error prone. To fully exploit the benefits of digital imagery, these processes should be automated. As a step toward this goal, we present an algorithm that extracts the structure of a parking lot visible from a given aerial image. To minimize human intervention in the use of aerial imagery, we devise a self-supervised learning algorithm that automatically generates a set of parking spot templates to learn the appearance of a parking lot and estimates the structure of the parking lot from the learned model. The data set extracted from a single image alone is too small to sufficiently learn an accurate parking spot model. However, strong priors trained using large data sets collected across multiple images dramatically improve performance. Our self-supervised approach outperforms the prior alone by adapting the distribution of examples toward that found in the current image. A thorough empirical analysis compares leading state-of-the-art learning techniques on this problem.

EUV Mask Inspection System

Abstract: Disclosed are apparatuses, methods, and lithographic systems for EUV mask inspection. An EUV mask inspection system can include an EUV illumination source, an optical system, and an image sensor. The EUV illumination source can be a standalone illumination system or integrated into the lithographic system, where the EUV illumination source can be configured to illuminate an EUV radiation beam onto a target portion of a mask. The optical system can be configured to receive at least a portion of a reflected EUV radiation beam from the target portion of the mask. Further, the image sensor can be configured to detect an aerial image corresponding to the portion of the reflected EUV radiation beam. The EUV mask inspection system can also include a data analysis device configured to analyze the aerial image for mask defects.

gRAID : A Geospatial Real-Time Aerial Image Display for a Low-Cost Autonomous Multispectral Remote Sensing

Abstract: gRAID: A Geospatial Real-Time Aerial Image Display for a Low-Cost Autonomous Multispectral Remote Sensing Platform (AggieAir)

Recognizing image environment from image and position

Abstract: A method of recognizing the environment of an image from an image and position information associated with the image includes acquiring the image and its associated position information; using the position information to acquire an aerial image correlated to the position information; identifying the environment of the image from the acquired aerial image; and storing the environment of the image in association with the image for subsequent use.

Method and apparatus for measuring of masks for the photo-lithography

Abstract: The invention relates to a method and an apparatus for measuring masks for photolithography. In this case, structures to be measured on the mask on a movable mask carrier are illuminated and imaged as an aerial image onto a detector, the illumination being set in a manner corresponding to the illumination in a photolithography scanner during a wafer exposure. A selection of positions at which the structures to be measured are situated on the mask is predetermined, and the positions on the mask in the selection are successively brought to the focus of an imaging optical system, where they are illuminated and in each case imaged as a magnified aerial image onto a detector, and the aerial images are subsequently stored. The structure properties of the structures are then analyzed by means of predetermined evaluation algorithms. The accuracy of the setting of the positions and of the determination of structure properties is increased in this case.

Fusion of LIDAR and aerial imagery for accurate building footprint extraction

Abstract: Building footprint extraction from GIS imagery/data has been shown to be extremely useful in various urban planning and modeling applications. Unfortunately, existing methods for creating these footprints are often highly manual and rely largely on architectural blueprints or skilled modelers. Although there has been quite a lot of research in this area, most of the resultant algorithms either remain unsuccessful or still require human intervention, thus making them infeasible for practical large-scale image processing systems. In this work, we present novel LiDAR and aerial image processing and fusion algorithms to achieve fully automated and highly accurate extraction of building footprint. The proposed algorithm starts with initial building footprint extraction from LiDAR point cloud based on an iterative morphological filtering approach. This initial segmentation result, while indicating locations of buildings with a reasonable accuracy, may however produce inaccurate building footprints due to the low resolution of the LiDAR data. As a refinement process, we fuse LiDAR data and the corresponding color aerial imagery to enhance the accuracy of building footprints. This is achieved by first generating a combined gradient surface and then applying the watershed algorithm initialized by the LiDAR segmentation to find ridge lines on the surface. The proposed algorithms for automated building footprint extraction have been implemented and tested using ten overlapping LiDAR and aerial image datasets, in which more than 300 buildings of various sizes and shape exist. The experimental results confirm the efficiency and effectiveness of our fully automated building footprint extraction algorithm.

Inverse lithography for 45-nm-node contact holes at 1.35 numerical aperture

Abstract: Inverse lithography technology (ILT) is a procedure that optimizes the mask layout to produce an image at the wafer with the targeted aerial image. For an illumination condition optimized for dense pitches, ILT inserts model-based subresolution assist features (AF) to improve the imaging of isolated features. ILT is ideal for random contact hole patterns, in which the AF are not at intuitive locations. The raw output of ILT consists of very complex smooth shapes that must be simplified for an acceptable mask write time. It is challenging for ILT to quickly converge to the ideal pattern as well as to simplify the pattern to one that can be manufactured quickly. ILT has many parameters that effect process latitude, background suppression, conversion run time, and mask write time. In this work, an optimization procedure is introduced to find the best tradeoff between image quality and run time or write time. A conversion run time reduction of 4.7× is realized with the outcome of this optimization procedure. Simulations of mask write time quantify the ability of ILT to be used for full chip applications. The optimization procedure is also applied to alternate mask technologies to reveal their advantages over commonly used 6% attenuated phase shift masks.

Aerial image segmentation for flood risk analysis

Abstract: This paper presents a technique for image segmentation. We demonstrate its efficacy for classsifying high-resolution aerial images. The application is peak water flow estimation in a river catchment in the city of Zurich and the data covers a large rural and urban setting. The output of the segmentation process is used as input to a hydrological model. We introduce a combined, probabilistic, segmentation approach based on colour (the LAB colour space is used), texture (using entropy) and image features (gradients). Classification rates for natural land surfaces and man-made structures are up to 90% and 85% respectively. When the automatic segmentation result is compared to the official land use data and reclassified for use in GIS we achieve an overall classification accuracy of 70%. This new classification is tested on the WetSpa hydrological model and the resulting flow estimate compares favourably with that computed from hand-classified land use data.

Orientation of Airborne Laser Scanning Point Clouds with Multi-View, Multi-Scale Image Blocks

Abstract: Comprehensive 3D modeling of our environment requires integration of terrestrial and airborne data, which is collected, preferably, using laser scanning and photogrammetric methods. However, integration of these multi-source data requires accurate relative orientations. In this article, two methods for solving relative orientation problems are presented. The first method includes registration by minimizing the distances between of an airborne laser point cloud and a 3D model. The 3D model was derived from photogrammetric measurements and terrestrial laser scanning points. The first method was used as a reference and for validation. Having completed registration in the object space, the relative orientation between images and laser point cloud is known. The second method utilizes an interactive orientation method between a multi-scale image block and a laser point cloud. The multi-scale image block includes both aerial and terrestrial images. Experiments with the multi-scale image block revealed that the accuracy of a relative orientation increased when more images were included in the block. The orientations of the first and second methods were compared. The comparison showed that correct rotations were the most difficult to detect accurately by using the interactive method. Because the interactive method forces laser scanning data to fit with the images, inaccurate rotations cause corresponding shifts to image positions. However, in a test case, in which the orientation differences included only shifts, the interactive method could solve the relative orientation of an aerial image and airborne laser scanning data repeatedly within a couple of centimeters.