The image segmentation method we use for leucocytes is based on image distance transformation, combining the region and edge approach, taking full advantage of image information. According to the shape, texture and color appearance of cells, we select 22 feature values and measure them. The classifier is designed on the statistical classification. A test for recognizing 831 leucocytes in 560 images shows that the classification accuracy is 96%. Clinical experts confirm this system; for it can automatically recognize leucocytes by pattern recognition technique, and it is demonstrably valid in practice.
Algorithms ; Artificial Intelligence ; Humans ; Leukocytes ; cytology ; Pattern Recognition, Automated ; methods

In this paper, an image mosaic algorithm based on feature points was proposed by making a study of the image mosaic methods and the characteristics of pathological image. Points of interest were firstly extracted by Moravec operator in this method, and the corresponding feature matches of the original images were achieved by correlation coefficient. The parameters of affine transformation were calculated by the matching feature points, and weighted average method was used to fuse images. The experimental results show that, as for the pathological image with translation and rotation, the algorithm can extract the matching feature points accurately and realize the image seamless mosaic.
Algorithms ; Image Enhancement ; methods ; Image Processing, Computer-Assisted ; methods ; Pathology ; methods ; Pattern Recognition, Automated ; methods

Discrimination of abnormal images from the numerous wireless capsule endoscope (WCE) video sequence images is laborious and time-consuming, so that a computer-based automatic image recognition system is desired for this task. We propose an algorithm to allow feature extraction from each image channel and decision fusion using multiple BP neural networks. The algorithm was tested and the results demonstrated its high efficiency and accuracy in identification of abnormalities in the WCE images.
Algorithms ; Capsule Endoscopy ; methods ; Image Interpretation, Computer-Assisted ; methods ; Pattern Recognition, Automated ; methods

At present, ray-casting algorithm is the most widely used algorithm in the field of medical image visualization, and it can achieve the best image quality. Due to large amounts of computation like sampling, gradient, lighting and blending calculation, the cost of ray-casting algorithm is very large. The characteristic of Streaming single instruction multiple datum extensions (SSE) instruction--supporting vector computation--can satisfy the property of ray-casting algorithm well. Therefore, in this paper, we improved the implementation efficient significantly by vectorization of gradient, lighting and blending calculation, and still achieved a high quality image at the same time.
Algorithms ; Humans ; Image Interpretation, Computer-Assisted ; methods ; Image Processing, Computer-Assisted ; methods ; Pattern Recognition, Automated ; methods

In this paper, some familiar algorithms of edge detection in ultrasound B-scan images are analyzed and studied. The results show that Sobel, Prewitt and Laplacian operators are sensitive to noise, Hough transform adapts to the whole detection, while LoG algorithm's average is zero and it couldn't change the whole dynamic area. Accordingly LoG algorithm is preferable.
Algorithms ; Artifacts ; Humans ; Image Enhancement ; methods ; Pattern Recognition, Automated ; methods ; Ultrasonography ; methods

Digital image processing is widely used in medical image. Segmentation is one of the important step in computer image processing. The essay introduces the procedure and effect of cell image segmentation by watershed algorithm.
Algorithms ; Cytological Techniques ; methods ; Image Processing, Computer-Assisted ; methods ; Pattern Recognition, Automated ; methods

OBJECTIVETo propose an improved adaptive spectral clustering method for image segmentation to allow automatic selection of the optimal scaling parameters and enhance the accuracy of spectral clustering.

METHODSUsing constrain conditions for optimizing the criterion function and determining the optimal scaling parameters by iteration, the final image segmentation was achieved through spectral clustering based on Nystrom approximation. We chose suit weight functions for different texture images, and used the proposed method for image segmentation. The k-means algorithm and the method of spectral clustering after pre-segmentation by manually choosing the scaling parameter were compared with the proposed method.

RESULTSThe improved spectral clustering algorithm with automatic selection of the optimal scaling parameters achieved better results of image segmentation than the other two methods.

CONCLUSIONThe proposed algorithm can improve the accuracy of spectral clustering for image segmentation.

The lip color of a person is closely related to his or her health in the visual diagnosis of traditional Chinese medicine (TCM). The traditional method to judge the color of lips is through observing by a TCM doctor. The diagnosis result is affected not only by the doctor's knowledge and diagnosis experience, but also by the light, temperature and other environmental impacts. For these reasons, sometimes different doctors may make different judgement for the same lips. So it is urgently needed that an objective evaluation as reference for doctors can be obtained. A method based on support vector machine (SVM) that classifies lip color by computer automatically is presented in the present paper. Firstly, nine features of lip color in Hue, Saturation and Intensity (HSI) color space were extracted. Then, according to different combinations of these features five different experiments were conducted. By comparing the results of these experiments, it was discovered that the mean value is one of the most important features for the lip color. The overall effect of classification is better when the mean value and variance of HSI were chosen than other characteristics. In addition, experiments results demonstrated that the accuracy rate of classification is not improved when more features were adopted. The objective of the present paper is to select the appropriate characteristics and to combine them effectively to classify lip colors.
Color ; Diagnosis, Differential ; Lip ; Medicine, Chinese Traditional ; methods ; Pattern Recognition, Automated ; methods ; Support Vector Machine

Microscopic characteristics of several Mongolian Herbal flowers were extracted by improved Pseudo-Jacobi (p = 4, q = 2)-Fourier Moments (PJFM's), and 368 different versions of 28 microscopic characteristics of these herbs were identified by using the minimum-mean-distance rule. The experimental results showed that the average identification rate reaches as high as 98.1%. Therefore, this study can provide new techniques for digitalization and visualization of microscopic characteristics of Mongolian Herbs.
China ; Flowers ; ultrastructure ; Image Processing, Computer-Assisted ; methods ; Pattern Recognition, Automated ; methods ; Plants, Medicinal ; ultrastructure

This paper presents a new method for automatic sleep stage classification which is based on the EEG permutation entropy. The EEG permutation entropy has notable distinction in each stage of sleep and manifests the trend of regular transforming. So it can be used as features of sleep EEG in each stage. Nearest neighbor is employed as the pattern recognition method to classify the stages of sleep. Experiments are conducted on 750 sleep EEG samples and the mean identification rate can be up to 79.6%.
Classification ; methods ; Electroencephalography ; methods ; Entropy ; Humans ; Pattern Recognition, Automated ; Signal Processing, Computer-Assisted ; Sleep Stages ; physiology