http://nova.newcastle.edu.au/vital/access/services/Feed ${session.getAttribute("locale")} 5 http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:12760 We present a method to automatically outline the optic disc in a retinal image. Our method for finding the optic disc is based on the properties of the optic disc using simple image processing algorithms which include thresholding, detection of object roundness and circle detection by Hough transformation. Our method is able to recognize the retinal images with general properties and the retinal images with variance of unusual properties since the parameters of our method can be flexibly changed by the unusual properties. 2013-04-16T00:50:16.743Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:6051 Due to the wide viewer-ship and high commercial potentials, recently, sports video analysis attracts extensive research efforts. One of the main tasks in sports video analysis is to identify sports genres i.e. sports video categorization. Most of the existing work focus on mapping content-based features to sports genres by using supervised learning methods. Moreover, video data sets seeks efficient data reduction methods due to the large size and noisy data. It lacks comparison analysis on the implementation and performance of these methods. In this paper, the research is carried out by using four dominant machine learning algorithms, namely Decision Tree, Support Vector Machine, K Nearest Neighbor and Naive Bayesian, and comparing their performance on a high dimensional feature set which selected by some feature selection tools such as Correlation-based Feature Selection (CFS), Principal Components Analysis (PCA) and Relief. Experimental results shows that Support Vector Machine (SVM) and k-NN are not sensitive to reduction of training sets. Moreover, three different feature reduction methods perform very differently with respect to four different tools. 2013-04-08T03:24:44.919Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:6212 This paper introduces an interactive and intelligent approach for accurate brain segmentation. A high resolution 3-Tesla magnetic resonance (MR) dataset was tested by state of the art automated algorithms as well as segmented by making use of the proposed interactive tools. The results show that the automated algorithms gave an incomplete or anatomically incorrect brain surface. About 4% false positive and 10% false negative error rates were reported by evaluating three automated methods. The proposed approach improved the quality and accuracy of the segmented results. 2013-04-08T03:23:20.211Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:6153 We present the segmentation of the foreground objects and the identification of the individual objects in the cigarette tin package, so the information will be used for the classification of the acceptable cases or defective cases. Visual inspection and classification of cigarette tin package are very important in manufacturing cigarette products that require high quality package. For the accurate automated inspection and classification, computer vision has been deployed widely in manufacturing. This paper concerned with the problem of identifying the individual cigarette in the tin packing using the image processing and morphology operations. The identified objects can be used for developing a defect finding system in the cigarette packing industries. The approach has two steps: (i) colour-based segmentation of the region of interests, (ii) identifying of individual object. The segmentation performance was evaluated on 18 images including the good cases and the defective cases. 2013-04-08T03:22:09.403Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:10116 This paper presents a novel and effective method of shape analysis and recognition based on skeleton and morphological structure. A series of preprocessing algorithms, smooth following and liberalization are introduced, and series of morphological structural points of image contour are extracted and merged. A series of basic shapes and a main shape of object image are described and segmented based on skeleton and morphological structure. Object shape is efficiently analyzed and recognized based on the extracted series of basic shapes and main shape. Comparing with other methods, the proposed method need not sample training set. Also, the new method can be used to analyze and recognize the shape structure of any shape, and there is no any requirement for the processed image data set. The new method can be used in image analysis, intelligent recognition, techniques, applications, systems and tools. 2013-04-07T22:37:12.209Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:10114 Cervical cancer is the second most common cancer among women. Meanwhile, cervical cancer could be largely preventable and curable with regular Pap tests. Nuclei changes in the cervix could be found by this test. Accurate nuclei detection is extremely critical as it is the previous step of analysing nuclei changes and diagnosis afterwards. Recently, computer-aided nuclei segmentation has increased dramatically. Although such algorithms could be utilised in the situation for sparse nuclei since they are intuitively detected, the segmentation for the complicated nuclei clusters is still challenging task. This paper presents a new methodology for the detection of cervical nuclei clusters. We first detect all the nuclei from the cervical microscopic image by an ellipse fitting algorithm. Second, we chose some high-relevant features from all the features we obtained in last step via F-score, which is based on to what extent one feature attributes to results. All the ellipses are then classified into single ones and cluster ones by C4.5 decision tree with selected features. We evaluated the performance of this method by the classification accuracy, sensitivity, and cluster predictive value. With the 9 selected features from the original 13 features, we came by the promising classification accuracy (97.8%). 2013-04-05T01:05:08.884Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:11563 Cervical cancer is the second most common malignancy among women worldwide, if it is detected in early stage, cure rate is relatively high. Computer aided abnormality detection for cervical smear is developed to assist medical experts to handle the microscopy images, examine cell abnormalities and diagnose dyskaryosis. The microscopy images of cells in cervix uteri are stained by the tumor marker Ki-67, so that the abnormal nuclei present brown while normal ones are bluish. Segmentation is the most important and difficult task to calculate the ratio of abnormal nuclei to all nuclei. In order to achieve accurate segmentation of nuclei, we propose a multi-level segmentation approach for abnormality identification in microscopy images. First level segmentation aims to partition abnormal (stained) nuclei regions and all nuclei regions. Because of under-segmentation after first level segmentation, second level segmentation is applied to further partition the clustered nuclei. In order to classify touching regions of clustered nuclei and separate regions of single nucleus, relevant meaningful features are extracted from regions of interest. Consequently all the nuclei regions are separated and in conjunction with the abnormal nuclei regions in the first level segmentation, the abnormality i.e. ratio of abnormal nuclei to all nuclei is obtained. Experimental results indicate that our method achieved an accuracy of 93.55% and 95.8% in term of abnormal nuclei and all nuclei respectively for identification of abnormalities. Our proposed method produces a satisfactory segmentation. 2013-04-05T01:02:52.642Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:10049 Packaging appearance is extremely important in cigarette manufacturing. Typically, there are two types of cigarette packaging defects: (1) cigarette laying defects such as incorrect cigarette numbers and irregular layout; (2) tin paper handle defects such as folded paper handles. In this paper, an automated vision-based defect inspection system is designed for cigarettes packaged in tin containers. The first type of defects is inspected by counting the number of cigarettes in a tin container. First k-means clustering is performed to segment cigarette regions. After noise filtering, valid cigarette regions are identified by estimating individual cigarette area using linear regression. The k clustering centers and area estimation function are learned off-line on training images. The second kind of defect is detected by checking the segmented paper handle region. Experimental results on 500 test images demonstrate the effectiveness of the proposed inspection system. The proposed method also contributes to the general detection and classification system such as identifying mitosis in early diagnosis of cervical cancer. 2013-04-05T00:58:55.101Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:12730 Intensity adjustment is an image enhancement technique that maps an image's intensity values to a new range. However, this intensity adjustment does not effectively enhance particular structures such as ribs on a 2D chest radiograph. In this paper, we develop a new method using a lopsided hemi-ellipsoid cavity to deflate lungs. This is necessary in order to enhance the unclear ribs resulting from air-filled lungs in a typical chest radiograph procedure. 2013-04-04T21:51:08.919Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:12729 Prediction of conversion from mild cognitive impairment (MCI) to Alzheimer's disease (AD) is of major interest in AD research. A large number of potential predictors have been proposed, with most investigations tending to examine one or a set of related predictors. In this study, we simultaneously examined multiple features from different modalities of data, including structural magnetic resonance imaging (MRI) morphometry, cerebrospinal fluid (CSF) biomarkers and neuropsychological and functional measures (NMs), to explore an optimal set of predictors of conversion from MCI to AD in an Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort. After FreeSurfer-derived MRI feature extraction, CSF and NM feature collection, feature selection was employed to choose optimal subsets of features from each modality. Support vector machine (SVM) classifiers were then trained on normal control (NC) and AD participants. Testing was conducted on MCIc (MCI individuals who have converted to AD within 24 months) and MCInc (MCI individuals who have not converted to AD within 24 months) groups. Classification results demonstrated that NMs outperformed CSF and MRI features. The combination of selected NM, MRI and CSF features attained an accuracy of 67.13%, a sensitivity of 96.43%, a specificity of 48.28%, and an AUC (area under curve) of 0.796. Analysis of the predictive values of MCIc who converted at different follow-up evaluations showed that the predictive values were significantly different between individuals who converted within 12 months and after 12 months. This study establishes meaningful multivariate predictors composed of selected NM, MRI and CSF measures which may be useful and practical for clinical diagnosis. 2013-04-03T03:24:27.257Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:8772 This paper presents an automatic liver parenchyma segmentation algorithm that can segment liver in abdominal CT images. There are three major steps in the proposed approach. Firstly, a texture analysis is applied to input abdominal CT images to extract pixel level features. In this step, wavelet coefficients are used as texture descriptors. Secondly, support vector machines (SVMs) are implemented to classify the data into pixel-wised liver area or non-liver area. Finally, integrated morphological operations are designed to remove noise and finally delineate the liver. Our unique contributions to liver segmentation are twofold: one is that it has been proved through experiments that wavelet features present good classification result when SVMs are used; the other is that the combination of morphological operations with the pixel-wised SVM classifier can delineate volumetric liver accurately. The algorithm can be used in an advanced computer-aided liver disease diagnosis and liver surgical planning system. Examples of applying the proposed algorithm on real CT data are presented with performance validation based on the comparison between the automatically segmented results and manually segmented ones. 2013-02-20T04:00:36.538Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:8730 A fully automated approach is presented to extract brain areas efficiently from FDG-PET head scans. A threshold value is automatically calculated from the histogram graph of the brain images, followed by region growing and morphological operations, to segment brain areas from these images. Next, the midsagittal lines on axial slices are detected to separate the brain into two hemispheres. The proposed approach has been applied to 226 cases of normal controls and patients with neurological diseases. The average processing time is about 3 seconds on a standard personal computer. The experiment has shown promising results. 2013-02-20T04:00:10.081Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:11801 Human-centred multimedia applications are a set of activities that human directly interact with multimedia, which consists of different forms. Within all multimedia, video is an ultimate resource, by which people could obtain sensory information. Since limitations on the capacity of imaging devices as well as shooting conditions, we cannot usually acquire high quality video records that desired. This problem could be addressed by super-resolution. We propose a novel scheme in the present paper for super-resolution problem, and make three contributions: (1) on the stage of image registration according to previous approaches, the reference image is picked out through observing or randomly. We utilise a simple but efficient method to select the base image; (2) a median-value image, rather than the average image used previously, is adopted as the initialization for estimate of super-resolution; (3) we adapt the traditional Cross Validation (CV) to a weighted version in the process of learning parameters from input observations. Experiments on synthetic and real data are provided to illustrate the effectiveness of our approach. 2012-10-24T03:12:33.504Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:11599 Cervical cancer is the second most common cancer among women. At the same time, cervical cancer could be largely preventable and curable with regular Pap tests. This test can find nuclei changes in the cervix. Accurate nuclei detection is extremely critical as it is the previous step of analysing nuclei changes and diagnosis afterwards. In recent years, automatic nuclei segmentation has increased dramatically. Although such algorithms could be utilised in the situation for sparse nuclei since they are intuitively detected, the segmentation for the complicated nuclei clusters is still challenging task. This paper presents a new methodology for the detection of cervical nuclei clusters. We first detect all the nuclei from the cervical microscopic image by an ellipse fitting algorithm. All the ellipses are then classified into single ones and cluster ones by C4.5 decision tree with elected features. We evaluated the performance of this method by the classification accuracy, sensitivity, and cluster predictive value. The result shown that the promising classification accuracy (97.8%) is obtained using C4.5 with 9 relative features. 2012-10-24T03:00:06.260Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:5876 Typically, inspiration is preferred when xraying the lungs. The x-ray technologist will ask a patient to be still and to take a deep breath and to hold it. This not only reduces the possibility of a blurred image but also enhances the quality of the image since air-filled lungs are easier to see on x-ray film. However, inspiration causes low density in the inner part of lung field. That means that ribs in the inner part of lung field have lower density than the other parts nearer to the border of the lung field. That is why edge detection algorithms often fail to detect ribs. Therefore to make rib edges clear we try to produce an expiration lung field using a 'hemi-elliptical cavity.' Based on the expiration lung field, we extract the rib edges using canny edge detector and a new connectivity method, called '4 way with 10-neighbors connectivity' to detect clavicle and rib edge candidates. Once the edge candidates are formed, our system selects the best candidates using knowledge-based constraints such as a gradient, length and location. The edges can be paired and labeled as superior rib edge and inferior rib edge. Then the system uses the clavicle, which is obtained in a same method for the rib edge detection, as a landmark to label all detected ribs. 2012-01-30T04:06:40.991Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:8831 Movie affective content detection attracts ever-increasing research efforts. However, the affective content analysis is still a challenging task due to the gap between low-level perceptual features and high-level human perception of the media. Moreover, clues from multiple modalities should be considered for affective analysis, since they were used in movies to represent emotions and romance emotional atmosphere. In this paper, mid-level representations are generated from low-level features. These mid-level representations are from multiple modalities and used for affective content inference. Besides video shots which is commonly used for video content analysis, audio sounds, dialogue and subtitle are explored to contribute to detect affective content. Since affective analysis rely on movie genres, experiments are implemented in respective genres. The results shows that audio sounds, dialogues and subtitles are effective and efficient for affective content detection. 2011-09-05T03:30:06.347Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:8830 The computer-assisted microscopy systems can increase the accuracy of the analysis. To guarantee correct results in computer-assisted microscopy, accurate nuclei segmentation is crucially important since images segmentation is the first step towards image understanding and image analysis. In this paper, we present clustering techniques to segment homogeneous clusters in RGB color space and then label each cluster as a different region. According to the evaluation process, 97% of nuclei pixels were correctly delineated with our algorithm and on average 90% of nuclei were correctly detected. Our methods could be of value to computer-based systems designed to objectively interpret microscopic images by accurate nuclei segmentation. 2011-09-05T03:20:04.030Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:8186 A fully automated approach is presented to extract brain efficiently from computed tomography (CT) head scans. Domain knowledge, including Hounsfield unit ranges, brain anatomy and image acquisition parameters, is applied. Regions of interest are first set in each slice by applying thresholding and region growing. Next, the brain candidates are extracted by using three-dimensional region growing with a variable, anatomy and acquisition-dependent structuring element. The proposed method has been applied automatically to 27 normal and pathological CT scans. The average processing time is four seconds for CT scans with 17–47 slices on a standard personal computer and the average sensitivity, specificity and Dice’s index for five cases are 99.6%, 99.4% and 98.7%, respectively. 2011-07-10T23:30:06.185Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:7793 Visual inspection and classification of cigarettes packaged in a tin container is very important in manufacturing cigarette products that require high quality package presentation. For accurate automated inspection and classification, computer vision has been deployed widely in manufacturing. We present the detection of the defective packaging of tins of cigarettes by identifying individual objects in the cigarette tins. Object identification information is used for the classification of the acceptable cases (correctly packaged tins) or defective cases (incorrectly packaged tins). This paper investigates the problem of identifying the individual cigarettes and a paper spoon in the packaged tin using image processing and morphology operations. The segmentation performance was evaluated on 500 images including examples of both good cases and defective cases. 2011-05-27T06:00:03.877Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:5878 A novel automatic system for extracting the semantic descriptions of medical image content and concept in text form is presented. We first extract and analyse image features and the features are mapped to semantic descriptions by fuzzy functions. Based on these semantic descriptions, our system facilitates knowledge base construction using a machine learning scheme. The result will be useful for other researchers in medical image retrieval area, who can take advantage of both text-based queries and image queries. 2010-07-20T02:40:08.507Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:2463 2010-04-27T06:24:10.675Z ]]> http://nova.newcastle.edu.au/vital/access/manager/Repository/uon:5938 This paper introduces an effective character extraction algorithm that can be used for optical character recognition (OCR). Using both geometrical and colour information, the character extraction algorithm can extract text from colour document images which contain mixed text and pictures. The algorithm consists of three components, i.e., adaptive k-means clustering, binary morphological processing, and shape and space-related refinement. When the algorithm is used as a plug-in pre-processing stage for an OCR system, the performance of the system can be improved. Character recognition experiment was done with a commercial OCR package. It has been shown that our algorithm can improve character recognition rate on complex document from 73.1% to 95.5% on average. 2010-04-27T04:32:01.021Z ]]>