Fuzzy control has been widely applied in industrial controls and domestic electrical equipment. The automatic learning of fuzzy rules is a key technique in fuzzy control. In this paper, a software development system f...
详细信息
Fuzzy control has been widely applied in industrial controls and domestic electrical equipment. The automatic learning of fuzzy rules is a key technique in fuzzy control. In this paper, a software development system for fuzzy control is presented. Since the learning of fuzzy rules can be seen as finding the best classifications of fuzzy memberships of input-output variables in a fuzzy controller, it can also be seen as the combination optimization on input-output fuzzy memberships. Multi-layer feedforward network and genetic algorithms (GA) can be used for the automatic learning of fuzzy rules. The algorithms and their characteristics are described. The software development system has been successfully used for the design of some fuzzy controllers.
K-means and watershed segmentation techniques are presented to perform image segmentation and edge detection tasks. We first used the K-means technique to obtain a primary segmented image. We then employed a watershed...
详细信息
K-means and watershed segmentation techniques are presented to perform image segmentation and edge detection tasks. We first used the K-means technique to obtain a primary segmented image. We then employed a watershed technique that works on that image;this process includes gradient of the segmented input image, divides the image into markers, completes the watershed line by using the markers, and stores the image in the format of region adjacency graph (RAG). The initial segmentation result was obtained by the watershed algorithm. We then used merging techniques based on mean gray values and two edge strengths (T1, T2) to obtain edge maps. In this article we solved the problem of undesirable oversegmentation results produced by the watershed algorithm, when used directly with raw data images. Also, the edge maps have no broken lines on the entire image, and the final edge detection result is one closed boundary per actual region in the image.
Due to the similarity in mushroom features and the difficulty in distinguishing between poisonous and nonpoisonous varieties, mushrooms pose a threat to human health. To address the challenge of mushroom classificatio...
详细信息
Simulating the soft tissue deformation based on bone-related planning is one of the most difficult issues in generating realistic motion of soft tissues. The difficulty of this problem arises from unclear boundary con...
详细信息
Segmentation and clustering of infrared small target images in a sky or sea-sky background is considered in this paper, which is the preprocessing part of the detection and recognition of the moving small targets in a...
详细信息
ISBN:
(纸本)0780384032
Segmentation and clustering of infrared small target images in a sky or sea-sky background is considered in this paper, which is the preprocessing part of the detection and recognition of the moving small targets in an infrared image sequence. The infrared image intensity surface is well fitted by the least squares support vector machines (LS-SVM), and then the maximum extremum points are detected on the well fitted intensity surface by convolving the image with the second order directional derivative operators deduced from the mapped LS-SVM with mixtures of kernels. With the coarse locations, the possible targets are extracted by the clustering analysis. The computer experiments are carried out for the real and simulated sky and sea-sky infrared images. The experimental results demonstrate the proposed approach is effective.
This thesis presents a circuit architecture to realize clock recovery for fast Ethernet application, including system architecture, modified Mueller Muller algorithm for 100BASE-TX, phase detector for 100BASE-FX and m...
详细信息
This thesis presents a circuit architecture to realize clock recovery for fast Ethernet application, including system architecture, modified Mueller Muller algorithm for 100BASE-TX, phase detector for 100BASE-FX and multiple output charge pump PLL. The clock generator circuit with 3.3 V operation voltage has been verified by TSMC 0.35 μm 1P5M CMOS process. The results show that this clock recovery circuit can exactly extract the timing information. It has advantages over other ones for simplicity and easy implementation.
We present an approach to detect anatomical structures by configurations of interest points, from a single example image. The representation of the configuration is based on Markov Random Fields, and the detection is ...
详细信息
ISBN:
(纸本)9783540757580
We present an approach to detect anatomical structures by configurations of interest points, from a single example image. The representation of the configuration is based on Markov Random Fields, and the detection is performed in a single iteration by the MAX-SUM algorithm. inst.ad of sequentially matching pairs of interest points, the method takes the entire set of points, their local descriptors and the spatial configuration into account to find an optimal mapping of modeled object to target image. The image information is captured by symmetry-based interest points and local descriptors derived from Gradient Vector Flow. Experimental results are reported for two data-sets showing the applicability to complex medical data.
To overcome the drawbacks of print measurements, a novel measurement approach, namely dot structure based measurement, is proposed. In addition, the structure feature extraction of microscopic dot is discussed. The pr...
详细信息
To overcome the drawbacks of print measurements, a novel measurement approach, namely dot structure based measurement, is proposed. In addition, the structure feature extraction of microscopic dot is discussed. The proposed scheme provides a precondition for such further researches as print quality surveillance, fault diagnosis, color design and printing theories as well.
This paper presents improvements in generation of wideband and high dynamic range analog signal for area-efficient MADBIST, especially for the on-chip testing of wireless communication IF digitizing sigma-delta modula...
详细信息
Fully-supervised learning requires expensive and laborious annotations of labeled data for crowd-counting tasks. To alleviate this burden, it is desirable to explore methods that reduce the need for extensive labeling...
Fully-supervised learning requires expensive and laborious annotations of labeled data for crowd-counting tasks. To alleviate this burden, it is desirable to explore methods that reduce the need for extensive labeling. Fortunately, there are a vast number of unlabeled images available in the world, making them easily accessible compared to labeled datasets. This paper proposes a self-supervised learning-based M-CNN framework with an attention mechanism that aims to leverage unlabeled data for pre-training the model. The framework consists of four sub-modules: a data augmentation framework, a self-supervised training network, a multi-column CNN, and an attention mechanism. These networks receive the images that undergo random processing using two defined augmentation transformations. Transformed images are then subjected to self-supervised learning and fed to a feature extraction network. FEN consists of M-CNN with five convolutional branches to extract features at a multi-scale level. These extracted features are then employed as an attention mechanism to focus on the head or shoulder location of people. To evaluate the effectiveness of our proposed model, experiments are conducted on two public datasets: ShanghaiTech Part A, Part B, and UCFQNRF. The experimental results demonstrate that our approach outperforms state-of-the-art semi-supervised methods, showcasing the effectiveness of our proposed approach in leveraging both unlabeled and limited labeled data for crowd counting tasks.
暂无评论