Statistical edge detection of concealed weapons using artificial neural networks

In this paper, we investigate the ability of humans to recognize objects using different types of edges. Edges arise in images because of several different physical phenomena, such as shadow boundaries, changes in material albedo or reflectance, changes to surface normals, and occlusion boundaries. By constructing synthetic photo realistic scenes, we control which edges are visible in a rendered image to investigate the relationship between human visual recognition and that edge type. We evaluate the information conveyed by each edge type through human studies on object recognition tasks. We find that edges related to surface normals and depth are the most informative edges, while texture and shadow edges can confuse recognition tasks. This work corroborates recent advances in practical vision systems where active sensors capture depth edges (e.g. Microsoft Kinect) as well as in edge detection where progress is being made towards finding object boundaries instead of just pixel gradients. Further, we evaluate seven standard and state-of-the-art edge detectors based on the types of edges they find by comparing the detected edges with known informative edges in the synthetic scene. We suggest that this evaluation method could lead to more informed metrics for gauging developments in edge detection, without requiring any human labeling. In summary, this work shows that human proficiency at object recognition is due to surface normal and depth edges and suggests that future research should focus on explicitly modeling edge types to increase the likelihood of finding informative edges.

Edge detection has beneficial applications in the fields such as machine vision, pattern recognition and biomedical imaging etc. Edge detection highlights high frequency components in the image. Edge detection is a challenging task. It becomes more arduous when it comes to noisy images. This study focuses on fuzzy logic based edge detection in smooth and noisy clinical images. The proposed method (in noisy images) employs a 3×3 mask guided by fuzzy rule set. Moreover, in case of smooth clinical images, an extra mask of contrast adjustment is integrated with edge detection mask to intensify the smooth images. The developed method was tested on noise-free, smooth and noisy images. The results were compared with other established edge detection techniques like Sobel, Prewitt, Laplacian of Gaussian (LOG), Roberts and Canny. When the developed edge detection technique was applied to a smooth clinical image of size 270×290 pixels having 24 dB ‘salt and pepper’ noise, it detected very few (22) false edge pixels, compared to Sobel (1931), Prewitt (2741), LOG (3102), Roberts (1451) and Canny (1045) false edge pixels. Therefore it is evident that the developed method offers improved solution to the edge detection problem in smooth and noisy clinical images.

Edge detection is an essential procedure in biomedical image processing. Note that several edge detection schemes based on ghost imaging (GI) have been proposed, while a practical application environment is not considered. In this paper, we first investigate the edge detection of target objects inside biological tissue in a GI system. The performances of four edge detection operators are compared in GI through biological tissue when refractive index turbulence is considered. The numerical results demonstrate that the Standard operator performs better in edge extraction regardless of the strength of refractive index fluctuations. Our results may provide potential applications of GI in the field of biomedical imaging.

Spiking Self-organizing Maps for Classification Problem

Creating high-quality training sets is the first step in designing robust classifiers. However, it is fairly difficult in practice when the data quality is questionable (data is heterogeneous, noisy and/or massively large). In this paper, we show how to apply a genetic algorithm for evolving training sets from data corpora, and exploit it for artificial neural networks (ANNs) alongside other state-of-the-art models. ANNs have been proved very successful in tackling a wide range of pattern recognition tasks. However, they suffer from several drawbacks, with selection of appropriate network topology and training sets being one of the most challenging in practice, especially when ANNs are trained using time-consuming back-propagation. Our experimental study (coupled with statistical tests), performed for both real-life and benchmark datasets, proved the applicability of a genetic algorithm to select training data for various classifiers which then generalize well to unseen data.

This paper presents a novel feature extraction method using the combination of the Coordinate Logic Filters (CLF) and Artificial Neural Networks (ANN) applied to 2D signals (Images). The method consists of image enhancement by histogram adaptive equalization technique, features extraction by modifying gray levels applying a nonlinear adaptive transformation function and edge detection by Coordinate Logic Filters (CLF), generation, clustering and labelling of suboptimal features vectors by Self Organizing Map (SOM) Neural Network. For the detection we applied Back Propagation Neural Network (BPNN). This method is tested to detect Microcalcifications (MCs) in Regions of Interest (ROI) from mammograms. The experiment results show that the proposed method can locate MCs in an efficient way, moreover the method promise interesting advances in Medical Industry.

Image edges are considered to be a type of important information that can be extracted by applying detectors with different methodologies. Edge detection is a major and important topic in computer vision and image processing. The technique of detecting edges in an image serves as a step for pre-processing many image processing algorithms, such as image enhancement, image segmentation, image tracking and encoding [10]. Edge detection is at the heart of all stages in image processing and object recognition. It is a process of identifying and locating abrupt breaks in the image, where the breaks represent changes in the intensity of the pixels that make up the boundaries of the objects in the image. There are a number of edge detection techniques, each of which is designed to be sensitive to certain types of edges. This article presents some of them, then analyzes the errors that may occur. The implementation is made on the software product MATLAB.

Budaya aksara Lampung pada saat ini sudah jarang digunakan sehingga terancam punah. Untuk melestarikan budaya tersebut, maka dilakukan penelitian mengenai huruf aksara Lampung secara digital dengan metode pengolahan citra dan pengenalan pola. Pada penelitian ini, terbagi menjadi dua tahap yaitu tahap pelatihan dan tahap pengujian. Pada tahap pelatihan dilakukan pengolahan citra dan pelatihan jaringan saraf tiruan (JST) backpropagation. Salah satu tahapan dalam pengolahan citra adalah deteksi tepi. Dalam menganalisis karakter huruf aksara Lampung menggunakan metode deteksi tepi untuk melihat pola garis huruf tersebut. Metode pengolahan citra yang digunakan dalam mendeteksi tepi citra adalah metode deteksi tepi Roberts dan metode deteksi tepi Sobel. Hasil pengolahan citra akan masuk ke tahap pelatihan JST. Bobot yang didapat dari pelatihan kemudian akan digunakan pada tahap pengujian. Metode Sobel dapat mengenali karakter aksara Lampung lebih baik, dimana hasil dari pengujian karakter aksara Lampung memberikan nilai presentase error sebesar 28.5% dengan metode deteksi Roberts dan 14.5% dengan metode deteksi Sobel.

Millimeter-wave holographic imaging techniques have recently been developed for personnel surveillance applications at airports and other high-security checkpoints. Millimeter- wave imaging is useful for this application since millimeter-waves easily pass through common clothing materials yet are reflected from the human body and any items concealed by clothing. This allows a high-resolution imaging system to form an image revealing items concealed on the person imaged. A prototype imaging system developed at Pacific Northwest National Laboratory uses a scanned linear array of millimeter-wave antennas to capture wideband millimeter-wave data in approximately one second. This data is then mathematically reconstructed to form a high- resolution 3D image of the person being scanned. Millimeter- wave imaging has been demonstrated to be effective for detecting concealed weapons on personnel. Another imaging technique which could be applied to the weapon detection problem is acoustic imaging. Like millimeter-waves, ultrasonic acoustic waves can also penetrate clothing, and can be used to form relatively high-resolution images which can reveal concealed weapons on personnel. Acoustic imaging results have been obtained using wideband holographic imaging techniques nearly identical to the imaging techniques used for millimeter-wave imaging. Preliminary imaging results at 50 kHz indicate that acoustic imaging can be used to penetrate some types of common clothing materials. Hard clothing materials, such as leather on vinyl, are essentially opaque to acoustic waves at 50 kHz. In this paper, millimeter-wave and acoustic wave imaging techniques are compared for their effectiveness and suitability in weapon detection imaging systems. Experimental results from both imaging modalities are shown.

T3D edge detection from a depth image is an important technique of 3D object recognition in preprocessing. There are three types of 3D edges in a depth image called jump, convex roof, and concave roof edges. Conventional 3D edge detection based on ring operators has been proposed. The conventional ring operator can detect three types of 3D edges by classifying the response of Fourier transforms. Since the conventional method needs to apply Fourier transforms to all pixels of a depth image, real-time processing cannot be done due to high computational cost. Therefore, this paper presents a fast and reliable method of detecting three types of 3D edges by using a decision tree. The decision tree is trained under supervised learning from numerous synthesized depth images and labels by capturing depth relations between candidate pixels and pixels on a ring operator to classify 3D edges. The experimental results revealed that the proposed method has 25 times faster than the conventional method. This paper also presents some examples of 3D line and 3D convex corner detection based on results obtained with the proposed method.

Image segmentation is an important problem in image processing that has a wide range of applications in medicine, biomedicine and other fields of science and engineering. During the non-learning-based approaches, the techniques based on the partial differential equations and calculus of variation have attracted a lot of attention and acquired many achievements. Among the variational models, the Chan-Vese variational segmentation is a well-known model to solve the image segmentation problem. The level set methods are highly accurate methods to solve this model, and they do not depend on the edges. However, the performance of these methods depends on the level set function and its initial boundary too much. In this paper, we propose automatic initial boundary generation methods based on the edge detectors: Sobel, Prewitt, Roberts and Canny. In the experiments, we prove that among the four proposed initial boundary generation methods, the method based on the Canny edge detector brings the highest performance for the segmentation method. By combining the proposed initial boundary generation method based on the Canny edge detector, we implement the Chan-Vese model to segment biomedical images. Experimental results indicate we obtain improved segmentation results and compare different edge detectors in terms of performance.

We propose a method to quantitatively evaluate image processing methods on how well they simulate human visual processing through psychological and statistical tests. First, we focus on edge detection, an image processing technique. To set all edge detection methods fairly, we collect the edge data detected by humans. Then, the signal processing parameters for each edge detection method are adjusted with the same optimizing method based on minimizing the error between the edges detected by the detector and by humans. Two psychological experiments are conducted, confirming whether the edge detection parameters are optimized and evaluating the performance of edge detection methods. The results indicate that the optimizing method functioned comparatively well. The effectiveness rank of edge detection methods is also obtained.

Biomedical imaging with emerging technology plays a vital role in medical diagnosis and treatment. MRI is one such biomedical imaging technology. The incidence of brain tumor in increasing all over the world. Accurate detection of size and location of brain tumor helps in right diagnosis and treatment. In any kind of image processing, image segmentation is a fundamental step. Edge detection is an important technique in the analysis of images, which plays a crucial role in detecting the contour of brain tumor. In this paper, an improved algorithm called the Luminance edge detection algorithm is proposed and compared with the existing edge detection algorithms like Prewitt, Sobel, and Canny for the MRI images of the brain. PSNR and SSIM were calculated to compare the image quality of different edge detection techniques. Results revealed that the PSNR and SSIM values were slightly higher for the images obtained by the Luminance edge detection algorithm when compared to the images obtained by Prewitt, Sobel, and Canny edge detection algorithm. It proves that the proposed algorithm is better than the other edge detection algorithms in producing a quality image of the MRI brain. Hence it can be concluded that the proposed algorithm was better in detecting the contour of brain and brain tumor than the other edge detection techniques.

Purpose: As multiple statistical tests are performed, the probability of committing type 1 error- accepting a false-positive result- increases. A number of statistical methods can correct for this. This study aims to comprehensively review the literature and assess the frequency at which multiple tests are performed without correction in pediatric orthopaedic research. Methods: Previously published methods were utilized to query the Institute for Scientific Information Web of Science for all …

This study aimed at characterization of rainfall dynamics in a hot arid region of Gujarat, India by employing time-series modeling techniques and sustainability approach. Five characteristics, i.e., normality, stationarity, homogeneity, presence/absence of trend, and persistence of 34-year (1980–2013) period annual rainfall time series of ten stations were identified/detected by applying multiple parametric and non-parametric statistical tests. Furthermore, the study involves novelty of proposing sustainability concept for evaluating rainfall time series and demonstrated the concept, for the first time, by identifying the most sustainable rainfall series following reliability (R y), resilience (R e), and vulnerability (V y) approach. Box–whisker plots, normal probability plots, and histograms indicated that the annual rainfall of Mandvi and Dayapar stations is relatively more positively skewed and non-normal compared with that of other stations, which is due to the presence of severe outlier and extreme. Results of Shapiro–Wilk test and Lilliefors test revealed that annual rainfall series of all stations significantly deviated from normal distribution. Two parametric t tests and the non-parametric Mann–Whitney test indicated significant non-stationarity in annual rainfall of Rapar station, where the rainfall was also found to be non-homogeneous based on the results of four parametric homogeneity tests. Four trend tests indicated significantly increasing rainfall trends at Rapar and Gandhidham stations. The autocorrelation analysis suggested the presence of persistence of statistically significant nature in rainfall series of Bhachau (3-year time lag), Mundra (1- and 9-year time lag), Nakhatrana (9-year time lag), and Rapar (3- and 4-year time lag). Results of sustainability approach indicated that annual rainfall of Mundra and Naliya stations (R y = 0.50 and 0.44; R e = 0.47 and 0.47; V y = 0.49 and 0.46, respectively) are the most sustainable and dependable compared with that of other stations. The highest values of sustainability index at Mundra (0.120) and Naliya (0.112) stations confirmed the earlier findings of R y–R e–V y approach. In general, annual rainfall of the study area is less reliable, less resilient, and moderately vulnerable, which emphasizes the need of developing suitable strategies for managing water resources of the area on sustainable basis. Finally, it is recommended that multiple statistical tests (at least two) should be used in time-series modeling for making reliable decisions. Moreover, methodology and findings of the sustainability concept in rainfall time series can easily be adopted in other arid regions of the world.