Image Projection Method Research Articles

Dose-Incorporated Deep Ensemble Learning for Improving Brain Metastasis Stereotactic Radiosurgery Outcome Prediction

Purpose/Objective(s)To develop a novel deep ensemble learning model for accurate prediction of brain metastasis(BM) local control outcomes following stereotactic radiosurgery(SRS). Materials/MethodsA total of 114 BMs from 82 patients were evaluated, including 26 BMs that developed biopsy-confirmed local failure post-SRS. The SRS spatial dose distribution(Dmap) of each BM was registered to the planning contrast-enhanced T1(T1-CE) MR. Axial slices of the Dmap, T1-CE, and PTV segmentation(PTVseg) intersecting the BM center were extracted within a fixed field-of-view determined by the V60% in Dmap. A spherical projection was implemented to transform planar image content onto a spherical surface using multiple projection centers, and the resultant T1-CE/Dmap/PTVseg projections were stacked as a 3-channel variable. Four VGG-19 deep encoders were utilized in an ensemble design, with each sub-model using a different spherical projection formula as input for BM outcome prediction. In each sub-model, clinical features after positional encoding were fused with VGG-19 deep features to generate logit results. The ensemble's outcome was synthesized from the four sub-model results via logistic regression. A total of 10 model versions with random validation sample assignments were trained to study model robustness. Performance was compared to 1) a single VGG-19 encoder; 2) an ensemble with T1-CE MRI as the sole image input after projections; and 3) an ensemble with the same image input design without clinical feature inclusion. ResultsThe ensemble model achieved an excellent AUCROC=0.89±0.02 with high sensitivity(0.82±0.05), specificity(0.84±0.11), and accuracy(0.84±0.08) results. This outperformed the MRI-only VGG-19 encoder (sensitivity:0.35±0.01, AUC:0.64±0.08), the MRI-only deep ensemble (sensitivity:0.60±0.09, AUC:0.68±0.06), and the 3-channel ensemble without clinical feature fusion (sensitivity:0.78±0.08, AUC:0.84±0.03). ConclusionFacilitated by the spherical image projection method, a deep ensemble model incorporating Dmap and clinical variables demonstrated an excellent performance in predicting BM post-SRS local failure. Our novel approach could improve other radiotherapy outcome models and warrants further evaluation.

New Chung-Li Ionosonde in Taiwan: System Description and Preliminary Results

In spite of being interrupted several times in its long history of operation since 1950, the routine observation of the ionosphere with various ionosondes installed at the Chung-Li ionosphere station in Taiwan has been achieved successively for more than seven decades. In this article, the system characteristics of the latest Chung-Li ionosonde and algorithm developed by National Central University for ionogram scaling and true height analysis, which started to routinely operate in 2020, are introduced. The new Chung-Li ionosonde is a pulse radar that transmits a train of short pulses with respective carrier frequencies between 2 and 30 MHz at a frequency separation of 50 kHz. The duration of an entire frequency sweep is 294.13 s, which is divided into 561 frequency channels. The 16-bit complementary code is employed to increase the signal-to-noise of the reflected echoes. The observational range is from 70 to 1221 km with a range resolution of 3.84 km. We developed an algorithm for the Chung-Li ionosonde to automatically scale the ionogram such that the true height profile of the ionospheric electron density can be retrieved. The observed traces of the ordinary wave (O-wave) and extraordinary wave (X-wave) displayed on the ionogram were first identified and separated by using 2-dimensional autocorrelation analysis combined with the image projection method. The true height analysis used stepwise regression. With the help of the International Reference Ionosphere (IRI) model and Quasi-Parabolic Segment (QPS) model, we carried out true height analysis to retrieve the ionospheric electron density profile based on the O-wave trace. An examination showed that the ionospheric parameters (i.e., foF2, h’F2) retrieved from the automatic scaling algorithm were essentially in good agreement with those obtained from manual scaling. The ionosonde-measured foF2 and hmF2 were also compared with the FORMOSAT-7 measurements made with the GPS radio occultation technique. The results show that the correlation coefficient, root mean squared deviation, and mean difference were, respectively, in ranges from 0.878 to 0.93, 0.73 to 1.06 MHz, and −0.43 to −0.26 MHz for foF2 and in ranges from 0.701 to 0.8, 22.39 to 28.45 km, and −9.28 to 11.06 km for hmF2.

The vertical image projection method for migration stretch compensation

True amplitude inversion is often carried out without taking into account migration distortions to the wavelet. Seismic migration leaves a dip-dependent effect on the wavelet that can cause significant inaccuracies in the inverted impedances obtained from conventional inversion approaches based on 1D vertical convolutional modeling. Neglecting this effect causes misleading inversion results and leakage of dipping noise and migration artifacts from the higher frequency bands to the lower frequencies. I have observed that, despite the dip dependency of this effect, low-dip and flat events may also suffer if they are contaminated with cross-cutting noise, steep migration artifacts, and smiles. I adopt an efficient, effective, and reversible data preconditioning approach that accounts for the dip dependency of the wavelet and is applied to migrated images prior to inversion. My proposed method consists of integrating data with respect to the total wavenumber followed by the differentiation with respect to the vertical wavenumber. This process is equivalent to applying a deterministic dip-consistent preconditioning that projects the data from the total wavenumber to the vertical wavenumber axis. This preconditioning can be applied to pre- and poststack data as well as to amplitude-variation-with-offset attributes such as intercept and gradient before inversion. The vertical image projection methodology that I adopt here reduces the impact of migration artifacts such as cross-cutting noise and migration smiles and improves inverted impedances in synthetic and real data examples. In particular, I indicate that neglecting the proposed preconditioning leads to anomalously higher impedance values along the steeply dipping structures.

Image projection method for vehicle speed estimation model in video system

Image projection method for vehicle speed estimation model in video system

Effective Statistical Fringe Removal Algorithm for High-Sensitivity Imaging of Ultracold Atoms

High-sensitivity imaging of ultracold atoms is often challenging when interference patterns are imprinted on the imaging light. Such image noises result in low signal-to-noise ratio and limit the capability to extract subtle physical quantities. Here we demonstrate an advanced fringe removal algorithm for absorption imaging of ultracold atoms, which efficiently suppresses unwanted fringe patterns using a small number of sample images without taking additional reference images. The protocol is based on an image decomposition and projection method with an extended image basis. We apply this scheme to raw absorption images of degenerate Fermi gases for the measurement of atomic density fluctuations and temperatures. The quantitative analysis shows that image noises can be efficiently removed with only tens of reference images, which manifests the efficiency of our protocol. Our algorithm would be of particular interest for the quantum emulation experiments in which several physical parameters need to be scanned within a limited time duration.

Visual Speech Recognition for Daily Indonesian Words Based on Combination of Double Difference and Image Projection Method

In visual speech recognition, there are two approaches to lip feature extraction, which are based on appearance and shape. The appearance-based approach is usually better, because it provides visual features that not only include lip structure but also the visibility of the teeth and tongue. However, the disadvantage of this approach is that it produces too many features. Integration of double difference and horizontal-vertical image projection, is part of the appearance approach, in addition to using image projection as dimensional reduction. In previous study, the method has succeeded in recognizing 5 daily indonesian words, with data in form of videos recorded inside the room. In this study, we used the same words with 5 additional new words recorded outside the room. MLP (Multi Layer Perceptron) and SVM (Support Vector Machine) are used as classifiers. The word recognition process is evaluated using 10-fold cross validation. The method tested reached 88.92% on classification accuracy and 0.9948 on AUC (Area Under ROC Curve).

Excimer Laser Three-Dimensional Micromachining Based on Image Projection and the Optical Diffraction Effect

An excimer laser three-dimensional (3D) micromachining system is proposed based on a mask image projection method and the optical diffraction effect. The effects of optical diffraction on the laser machining rate are evaluated using a hole-arrayed mask pattern with various feature sizes and hole-area opening ratios. The practical feasibility of the proposed method is demonstrated by machining conical, trihedral, and pyramidal 3D microstructures on polycarbonate substrates. The proposed method greatly simplifies the photo-mask design and preparation task in traditional excimer laser 3D micromachining systems and provides a powerful technique for achieving large-area 3D microstructures with complex patterns and atypical profiles.

The automatic interpretation of structural plane parameters in borehole camera images from drilling engineering

Digital panoramic borehole camera technology has been widely employed in actual projects, and a large number of high-accuracy borehole camera images have been obtained. Borehole camera images accurately record the geological information, especially the feature parameters of structure plane. However, since the acquisition of these features is usually done by hand, the workload is large and the results can be affected by human error. To solve this problem, this paper presents an automatic interpretation method of structure plane parameters in borehole camera image. In this method, image pixel gray, gradient values and projection method are employed to distinguish the occurrence region of structural planes. Then, standard sine function matching method is employed to search the structure plane in the region. Lastly, the optimal sine curve is screened out and adopted as the feature curve of structure plane. Related parameters of sine curve are analyzed and converted into the parameters of structure plane, such as the central position, orientation, dip angle and fracture width of structural planes. This method can automatically identify structure planes in the image continuously and quickly, and obtain the corresponding structural parameters. The method is stable and reliable, and greatly improves the working efficiency. It can perform the automatic interpretation of structure planes and extraction of geometric parameters, and provide an effective and consistent solution for drilling information acquisition and borehole camera image signal processing.

簡易没入型ディスプレイCC Roomの開発と映像生成手法(「グラフィックスとアルゴリズム」特集)

簡易没入型ディスプレイCC Roomの開発と映像生成手法(「グラフィックスとアルゴリズム」特集)

A micro cutter auto-alignment system with on-machine positioning error measurement and compensation methods

To perform an accurate micro machining process, accurate alignment of the micro cutter with respect to the local coordinate frame for machining is essential. Due to the fragility of the micro tool, traditional cutter alignment method used in industry is no longer proper for micro machining. Utilizing machine vision technology with 2 CCDs, a micro tool positioning error measurement and auto-compensation method and system were proposed. In the measurement method, Power Law Method was first used to enhance the image taken by a CCD. Canny Edge Method and image projection method were then used to identify the contours of the cutter and the workpiece from the image. Finally, the position errors of the cutter were determined through calculating the number of pixels between the cutter and the workpiece. The proposed error compensation system can automatically generate NC codes which will accurately re-locate the cutter based on the measured position errors. Experiments were conducted on a micro machine tool, and the results have shown that the y-dir. positioning error of the cutter was significantly improved from −0.025 mm to 0.001 mm, and the total machining error was reduced from −0.02 mm to −0.001 mm.

Simple, all-optical, noncontact, depth-selective, narrowband surface acoustic wave measurement system for evaluating the Rayleigh velocity of small samples or areas

In this paper a new ultrasonic testing system is described that utilizes noncontact optical methods to generate and detect surface acoustic waves (SAWs) and has significant applications in the nondestructive evaluation of surface material. A narrowband SAW is generated with a new and straightforward grating mask image projection method that provides fast switching and a controllable frequency band, and hence control of the penetration depth of the ultrasonic wave. A narrowband SAW with center frequency above 30 MHz, and hence better depth resolution, is generated. The detection of the SAW is performed with a simplified design of an optical fiber interferometer that has good sensitivity and manoeuvrability without requiring additional auxiliary components. The novel combination of these two optical techniques permits the measurement of small samples that are otherwise difficult to measure, especially nondestructively. A model was constructed to simulate the temporal characteristics of the generated narrowband SAW and showed good agreement with experiment. Measurements on an aluminum sample and an extracted human incisor demonstrate the system's performance.

2123 金属ガラスの超塑性鍛造によるマイクロインダクターの開発(S29-1 マイクロ/ナノ加工の動向と可能性(1),S29 マイクロ/ナノ加工の動向と可能性)

We designed and fabricated a micro-inductor that may be used for hybrid electric circuits such as those in a mobile phone. The component was fabricated by superplastic die-forging of Pt-based metallic glass. This material exhibits Newtonian viscous flow in the supercooled liquid temperature range between the glass transition and the crystallization point. The working temperature was 543K, so a polyimide film was selected as a die material due to its good mechanical properties at elevated temperatures and its precision machinability. The die was fabricated with an excimer laser micro-machining system using the image-projection method. Laser ablation machining was carried out using a photomask in the shape of the micro-inductor. The photomask was fabricated using a lithography system. Superplastic die-forging of the metallic glass was carried out using this die, and micro-structures of the micro-inductor were easily fabricated. These fabrication processes demonstrate the advantages of metallic glasses and superplastic microforming techniques for mass producing micro/nano machine parts or devices.

Development of 2-dimensional Profile Automatic Measuring Machine Using Pinhole Image Projection Method

Development of 2-dimensional Profile Automatic Measuring Machine Using Pinhole Image Projection Method

論文小特集 ３次元映像情報メディア技術 曲面スクリーンを用いたマルチプロジェクションディスプレイにおける任意視点での歪みのない映像提示手法

Various shaped screens have bee recently used for immersive projection displays. The ability to project a non-distorted image on a curved screen from arbitrary viewpoints is highly desirable. However, in previous techniques, the shape of the screen has been restricted, and we required geometrical information about screens and projectors. In this paper, we descrive a non-distorted image projection method for an immersive projection display with a curved screen. This method contains two approaches : the first is a precise geometry and color correction for preprocessing with a digital camera, and the second is a real-time geometry correction with real-time 3D rendering techniques. These methods need only geometrical information about the screen, and achieve a high quality image correction in real-time. We also used this method in the actual immersive projection display, and evaluated its effectiveness in some experiments.

Recognition of English Calling Cards by Using Projection Method and Enhanced RBE Network

In this paper, we proposed the novel method for the recognition of English calling cards by using the projection method and the enhanced RBF (Radial Basis Function) network. The recognition of calling cards consists of the extraction phase of character areas and the recognition phase of extracted characters. In the extraction phase, first of all, noises are removed from the images of calling cards, and the feature areas including character strings are separated from the calling card images by using the horizontal smearing method and the 8-directional contour tracking method. And using the image projection method, the feature areas are split into the areas of individual characters. We also proposed the enhanced RBF network that organizes the middle layer effectively by using the enhanced ART1 neural network adjusting the vigilance threshold dynamically according to the homogeneity between patterns. In the recognition phase, the proposed neural network is applied to recognize individual characters. Our experiment result showed that the proposed recognition algorithm has higher success rate of recognition and faster learning time than the existing neural network based recognition.

AUTOMATED INSPECTION OF IC PRINTED MARK DEFECTS (II)—DEFECT DETECTION

ABSTRACT The study develops a defect detection system for the IC printed marks. The system can be divided into two main parts, namely, image pre-processing and defect detection. Image pre-processing includes binary image operation, image rotation, image segmentation and thinning operation. Defect detection includes character searching, characteristic extraction, characteristic comparison, image mapping and defect identification. As for the performance of this system, it can recognize the seven kinds of most common IC printed mark defects. For a case of 15mm and 8mm IC chip with 16 printed marks, the recognition speed averages about 0.54 seconds for each chip, and the accuracy is 99.5% for distinguishing a good/bad chip and 98.25% for distinguishing the defect types. In addition, the study proposes a two-stage character searching method, which is quicker than the image projection and segmentation method. Moreover, when the defect types do not need to be distinguished, the research suggests the character searching method coupling with the normalized cross correlation scheme; however, when distinguishing the defect types is necessary, the study proposes the characteristic comparison method. Both have the better results.