Segmentation Error Rate Research Articles

Feature extraction from smartphone inertial signals for human activity segmentation

This paper proposes the adaptation of well-known strategies successfully used in speech processing: Mel Frequency Cepstral Coefficients (MFCCs) and Perceptual Linear Prediction (PLP) coefficients. Additionally characteristics like RASTA filtering or delta coefficients are also considered and evaluated for inertial signal processing. These adaptations have been incorporated into a Human Activity Recognition and Segmentation (HARS) system based on Hidden Markov Models (HMMs) for recognizing and segmenting six different physical activities: walking, walking–upstairs, walking-downstairs, sitting, standing and lying.All experiments have been done using a publicly available dataset named UCI Human Activity Recognition Using Smartphones, which includes several sessions with physical activity sequences from 30 volunteers. This dataset has been randomly divided into six subsets for performing a six-fold cross validation procedure. For every experiment, average values from the six-fold cross-validation procedure are shown.The results presented in this paper overcome significantly baseline error rates, constituting a relevant contribution in the field. Adapted MFCC and PLP coefficients improve human activity recognition and segmentation accuracies while reducing feature vector size considerably. RASTA-filtering and delta coefficients contribute significantly to reduce the segmentation error rate obtaining the best results: an Activity Segmentation Error Rate lower than 0.5%.

Segmentation of foreground apple targets by fusing visual attention mechanism and growth rules of seed points

Accurate segmentation of apple targets is one of the most important problems to be solved in the vision system of apple picking robots. This work aimed to solve the difficulties that background targets often bring to foreground targets segmentation, by fusing the visual attention mechanism and the growth rule of seed points. Background targets could be eliminated by extracting the ROI (region of interest) of apple targets; the ROI was roughly segmented on the HSV color space, and then each of the pixels was used as a seed growing point. The growth rule of the seed points was adopted to obtain the whole area of apple targets from seed growing points. The proposed method was tested with 20 images captured in a natural scene, including 54 foreground apple targets and approximately 84 background apple targets. Experimental results showed that the proposed method can remove background targets and focus on foreground targets, while the k-means algorithm and the chromatic aberration algorithm cannot. Additionally, its average segmentation error rate was 13.23%, which is 2.71% higher than that of the k-means algorithm and 2.95% lower than that of the chromatic aberration algorithm. In conclusion, the proposed method contributes to the vision system of apple-picking robots to locate foreground apple targets quickly and accurately under a natural scene

Blurred trace infrared image segmentation based on template approach and immune factor

This paper proposes an image segmentation method for blurred trace infrared images, exploring statistical properties of blurred infrared data to segment target regions. I consider the function of immune factors for blurred infrared image segmentation, combined to the template algorithm framework. First, all of the pixel antigens are divided into three classes by innate recognition factors: target antigens, background antigens and blurred antigens. Next, the innate presentation factors present template characteristics as new features for each antigen. Finally, target and background antigens are used for generating mature adaptive immune factors, and these mature factors will recognise each blurred antigen into two classes: a target pixel or a background pixel. Experimental results indicate that the superstring galaxy template algorithm can improve the target segmentation rate and reduce the segmentation error rate.

Retinal Thickness and Volume Measured With Enhanced Depth Imaging Optical Coherence Tomography

To evaluate the retinal thickness and volume measured with the enhanced depth imaging (EDI) method compared with those measured with the conventional method using spectral-domain optical coherence tomography (OCT). Retrospective, observational, case-control study. Clinical records of 20 healthy subjects and those of 35 patients with chorioretinopathy (central serous chorioretinopathy, polypoidal choroidal vasculopathy, Vogt-Koyanagi-Harada disease, and reticular pseudodrusen) were analyzed retrospectively. All subjects underwent spectral-domain OCT using both the conventional and the EDI OCT raster scan protocols. The raster scan was composed of 31 B-scans that were 9.0 mm in length and 240 μm apart. Retinal thickness and volume of 9 Early Treatment Diabetic Retinopathy Study subfields were investigated. Intraclass correlation coefficients, Bland-Altman plots, and Wilcoxon signed-rank test results were used for the analysis. Sixty-five eyes of 35 patients with chorioretinal diseases and 40 eyes of 20 normal healthy subjects were evaluated. The automatically measured retinal thickness and volume of 9 Early Treatment Diabetic Retinopathy Study subfields with conventional and EDI raster scan showed an intraclass correlation coefficient of 0.861 to 0.995 and 0.873 to 0.995, respectively. The 95% limits of agreement between the 2 protocols in the measurement of central subfield were -14.52 to 12.88 μm in retinal thickness and -0.014 to 0.013 mm³ in retinal volume. The differences of segmentation error rate between the 2 protocols were statistically insignificant (P > .05), except in eyes with reticular pseudodrusen in the subgroup analysis (P = .006). No significant differences were observed in measured values between healthy eyes and unaffected fellow eyes. The EDI OCT raster scan showed high agreement with conventional OCT in the measurement of retinal thickness and volume and could be used to evaluate both the retina and choroid in normal eyes and in eyes with some forms of chorioretinal disorder.

Two-Level Evaluation on Sensor Interoperability of Features in Fingerprint Image Segmentation

Features used in fingerprint segmentation significantly affect the segmentation performance. Various features exhibit different discriminating abilities on fingerprint images derived from different sensors. One feature which has better discriminating ability on images derived from a certain sensor may not adapt to segment images derived from other sensors. This degrades the segmentation performance. This paper empirically analyzes the sensor interoperability problem of segmentation feature, which refers to the feature’s ability to adapt to the raw fingerprints captured by different sensors. To address this issue, this paper presents a two-level feature evaluation method, including the first level feature evaluation based on segmentation error rate and the second level feature evaluation based on decision tree. The proposed method is performed on a number of fingerprint databases which are obtained from various sensors. Experimental results show that the proposed method can effectively evaluate the sensor interoperability of features, and the features with good evaluation results acquire better segmentation accuracies of images originating from different sensors.

Stratus OCT image analysis with spectral-domain OCT (Topcon 3D OCT Viewer)

AimsTo assess the accuracy of the Topcon 3D OCT Viewer in analysing Stratus OCT data.MethodsRandomly selected fast macular thickness map protocol optical coherence tomography (OCT) raw data were exported from...

Comparison of retinal thickness values and segmentation performance of different OCT devices in acute branch retinal vein occlusion

To compare retinal thickness (RT) measurement and segmentation performance of time domain (TD, Stratus) and spectral domain (SD) optical coherence tomography (OCT) devices (Cirrus, Spectralis) for imaging macular oedema (ME) secondary to branch retinal vein occlusion (BRVO). In this study, 20 eyes of 20 consecutive patients with acute BRVO were included. A total of 18 unaffected fellow eyes served as control group. RT measurement was analysed in the five inner fields of the early-treatment diabetic retinopathy grid, and proportional segmentation errors were evaluated. Central millimetre thickness (CMT) showed a mean difference of -64, -74, and -18 μm (P < 0.001) in the control group and -31 μm (P=0.107), -92 μm (P<0.001), and -105 μm (P=0.016) in the BRVO group, between Stratus and Cirrus, between Stratus and Spectralis, and between Cirrus and Spectralis, respectively. Mean RT showed the highest variability between different devices in the area most intensively affected by BRVO-related ME. In eyes with BRVO, 14.6% of Spectralis, 20% of Stratus, and 36.6% of Cirrus scans demonstrated moderate and severe segmentation errors. RT measurement in eyes with BRVO, by TD and SD OCT, is compromised by a significant rate of segmentation errors. Deviations are most pronounced in the areas most severely affected by ME.

Comparison of Macular Thickness Measurement and Segmentation Error Rate Between Stratus and Fourier-Domain Optical Coherence Tomography

To compare macular thickness measurements and segmentation error rates between Stratus optical coherence tomography (OCT) (Carl Zeiss Meditec, Inc., Dublin, CA), and Fourier-domain OCT (RTVue, Optovue, Inc., Fremont, CA). A retrospective study was performed of 93 normal and pathologic eyes from 79 subjects imaged with both OCT instruments on the same day. Both the macular thickness measurement for each Early Treatment Diabetic Retinopathy Study (ETDRS) zone and the incidence of segmentation error in the central macula between the two instruments were compared. Macular thickness measurements for all nine ETDRS zones were higher with RTVue compared with Stratus OCT (P < .01). Linear regression analysis showed the highest correlation in the central macula (R(2) = 0.88), with progressively lower correlation peripherally. The overall segmentation error rate was 29% with Stratus OCT versus 32% with RTVue (P > .05). Macular thickness measurement was greater with RTVue than with Stratus OCT in all ETDRS areas, with the best correlation seen in the central macula. No difference in segmentation error rate was noted between the two instruments.

A Robust Hidden Markov Gauss Mixture Vector Quantizer for a Noisy Source

Noise is ubiquitous in real life and changes image acquisition, communication, and processing characteristics in an uncontrolled manner. Gaussian noise and Salt and Pepper noise, in particular, are prevalent in noisy communication channels, camera and scanner sensors, and medical MRI images. It is not unusual for highly sophisticated image processing algorithms developed for clean images to malfunction when used on noisy images. For example, hidden Markov Gauss mixture models (HMGMM) have been shown to perform well in image segmentation applications, but they are quite sensitive to image noise. We propose a modified HMGMM procedure specifically designed to improve performance in the presence of noise. The key feature of the proposed procedure is the adjustment of covariance matrices in Gauss mixture vector quantizer codebooks to minimize an overall minimum discrimination information distortion (MDI). In adjusting covariance matrices, we expand or shrink their elements based on the noisy image. While most results reported in the literature assume a particular noise type, we propose a framework without assuming particular noise characteristics. Without denoising the corrupted source, we apply our method directly to the segmentation of noisy sources. We apply the proposed procedure to the segmentation of aerial images with Salt and Pepper noise and with independent Gaussian noise, and we compare our results with those of the median filter restoration method and the blind deconvolution-based method, respectively. We show that our procedure has better performance than image restoration-based techniques and closely matches to the performance of HMGMM for clean images in terms of both visual segmentation results and error rate.

Segmentation of optical coherence tomography images for differentiation of the cavernous nerves from the prostate gland

The cavernous nerves course along the surface of the prostate and are responsible for erectile function. Improvements in identification, imaging, and visualization of the cavernous nerves during prostate cancer surgery may improve nerve preservation and postoperative sexual potency. Two-dimensional (2-D) optical coherence tomography (OCT) images of the rat prostate were segmented to differentiate the cavernous nerves from the prostate gland. To detect these nerves, three image features were employed: Gabor filter, Daubechies wavelet, and Laws filter. The Gabor feature was applied with different standard deviations in the x and y directions. In the Daubechies wavelet feature, an 8-tap Daubechies orthonormal wavelet was implemented, and the low-pass sub-band was chosen as the filtered image. Last, Laws feature extraction was applied to the images. The features were segmented using a nearest-neighbor classifier. N-ary morphological postprocessing was used to remove small voids. The cavernous nerves were differentiated from the prostate gland with a segmentation error rate of only 0.058+/-0.019. This algorithm may be useful for implementation in clinical endoscopic OCT systems currently being studied for potential intraoperative diagnostic use in laparoscopic and robotic nerve-sparing prostate cancer surgery.

Arabic diacritic restoration approach based on maximum entropy models

In modern standard Arabic and in dialectal Arabic texts, short vowels and other diacritics are omitted. Exceptions are made for important political and religious texts and in scripts for beginning students of Arabic. Scripts without diacritics have considerable ambiguity because many words with different diacritic patterns appear identical in a diacritic-less setting. In this paper we present a maximum entropy approach for restoring short vowels and other diacritics in an Arabic document. The approach can easily integrate and make effective use of diverse types of information; the model we propose integrates a wide array of lexical, segment-based and part- of- speech tag features. The combination of these feature types leads to a high-performance diacritic restoration model. Using a publicly available corpus (LDC’s Arabic Treebank Part 3), we achieve a diacritic error rate of 5.1%, a segment error rate 8.5%, and a word error rate of 17.3%. In case-ending-less setting, we obtain a diacritic error rate of 2.2%, a segment error rate of 4.0%, and a word error rate of 7.2%. We also show in this paper a comparison of our approach to previously published techniques and we demonstrate the effectiveness of this technique in restoring diacritics in different kind of data such as the dialectal Iraqi Arabic scripts.

A statistical information extraction system for Turkish

This paper presents the results of a study on information extraction from unrestricted Turkish text using statistical language processing methods. In languages like English, there is a very small number of possible word forms with a given root word. However, languages like Turkish have very productive agglutinative morphology. Thus, it is an issue to build statistical models for specific tasks using the surface forms of the words, mainly because of the data sparseness problem. In order to alleviate this problem, we used additional syntactic information, i.e. the morphological structure of the words. We have successfully applied statistical methods using both the lexical and morphological information to sentence segmentation, topic segmentation, and name tagging tasks. For sentence segmentation, we have modeled the final inflectional groups of the words and combined it with the lexical model, and decreased the error rate to 4.34%, which is 21% better than the result obtained using only the surface forms of the words. For topic segmentation, stems of the words (especially nouns) have been found to be more effective than using the surface forms of the words and we have achieved 10.90% segmentation error rate on our test set according to the weighted TDT-2 segmentation cost metric. This is 32% better than the word-based baseline model. For name tagging, we used four different information sources to model names. Our first information source is based on the surface forms of the words. Then we combined the contextual cues with the lexical model, and obtained some improvement. After this, we modeled the morphological analyses of the words, and finally we modeled the tag sequence, and reached an F-Measure of 91.56%, according to the MUC evaluation criteria. Our results are important in the sense that, using linguistic information, i.e. morphological analyses of the words, and a corpus large enough to train a statistical model significantly improves these basic information extraction tasks for Turkish.

High resolution analysis of cervical cells--a progress report.

This paper presents preliminary results of research toward the development of a high resolution analysis stage for a dual resolution image processing-based prescreening device for cervical cytology. Experiments using both manual and automatic methods for cell segmentation are described. In both cases, 1500 cervical cells were analyzed and classified as normal or abnormal (dysplastic or malignant) using a minimum Mahalanobis distance classifier with eight subclasses of normal cells, and five subclasses of abnormal cells. With manual segmentation, false positive and false negative error rates of 2.98 and 7.73% were obtained. Similar experiments using automatic cell segmentation methods yielded false positive and false negative error rates of 3.90 and 11.56%, respectively. In both cases, independent training and testing data were used.