An automatic system to recognize the biological zooplanktons from monocular images is presented. It has two cooperating subsystems: a modular set of algorithm describing the complex shapes and a classification expert ...
详细信息
An automatic system to recognize the biological zooplanktons from monocular images is presented. It has two cooperating subsystems: a modular set of algorithm describing the complex shapes and a classification expert system. the results are shown for various examples. the methodology which is developed is independent of the application and is thus a general tool for building complex automatic systems.< >
Surface curvatures have been widely used for the purpose of segmenting and classifying range images. However, since surface curvatures include the second order derivatives, they become unreliable in the presence of no...
详细信息
A token matching scheme for high-level motion analysis has been developed. the goal is real-time usage for-a constrained scenario. the demand for real-time has forced an unsophisticated method, which simply propagates...
详细信息
Entropy of order q (depending on the information contained in a sequence of gray levels of length q) and conditional entropy of an image are defined. Using these definitions, two algorithms are formulated and implemen...
详细信息
Entropy of order q (depending on the information contained in a sequence of gray levels of length q) and conditional entropy of an image are defined. Using these definitions, two algorithms are formulated and implemented withthe help of its co-occurrence matrix. their superiority for imagethresholding (object-background classification) is established.< >
MIRACLE-IV is discussed, which is capable of obtaining an internal structure of an object from a series of silhouette images with no initial explicit models about the object. the images are derived from only one objec...
详细信息
MIRACLE-IV is discussed, which is capable of obtaining an internal structure of an object from a series of silhouette images with no initial explicit models about the object. the images are derived from only one object, but the form is varied. the system is composed of two subsystems: a model acquisition part (the modeler) and an imageprocessing strategy part (the strategist). On the assumption that the object consists of hinges, slides, and solids, the modeler learns their number in the object and the relationship among them. the strategist binds the functional features as hinges or slides with visual features in the actual image data. the imageprocessing sequence for the extraction of the visual feature is not given but is learned automatically through trial and error.< >
Techniques of image matching for stereo vision, generated withthe aid of a structured light projector, are described, and are used to match primitives. Based on the description function of the primitive, the match ru...
详细信息
Techniques of image matching for stereo vision, generated withthe aid of a structured light projector, are described, and are used to match primitives. Based on the description function of the primitive, the match rule is designed and four match constraints are used for eliminating the false match. A global to local control strategy is designed to accelerate the match process. this technique is used to obtain 3D measurements of human body, and to establish the surface representation for a computer-aided garment-design system. the results obtained indicate that this technique could also be used to acquire the 3D information for industrial parts and natural objects with smooth surfaces.< >
A method is described for the recognition of partially occluded 2-D objects. this method considers a set of corners, parallel lines, and so on as typical features of an object. Possible candidate models are estimated ...
详细信息
A method is described for the recognition of partially occluded 2-D objects. this method considers a set of corners, parallel lines, and so on as typical features of an object. Possible candidate models are estimated from these features, and structural matching is performed between these models and features obtained from a picture by checking the combinations of various features. Even if the whole structure is not obtained due to a partial occlusion, the system can infer an object if some unique features of the object are obtained. Partial shapes and extracted lines are matched in detail with model candidates when they are limited to one or a few.< >
A feature detector is described that bases its research on a set of probability distributions that describe the situations in which the feature can be found. the result is a measure of the probability of the feature o...
详细信息
A feature detector is described that bases its research on a set of probability distributions that describe the situations in which the feature can be found. the result is a measure of the probability of the feature occurring at a particular point. the probabilities required to operate this detector can be automatically derived from test images, although it can be altered to detect any kind of feature. the output from this detector is a set of probabilities that represent the likelihood of and edge occurring at each point in the image. these need only to be summed to find the probability of an edge over an interval, and, in a similar fashion, images can be subsampled in order to find probable edges to subpixel accuracy.< >
An algorithm for fitting a conic shape to a set of given data is described. By a modified medial axis transform, described in the paper, the algorithm first determines the orientation of the shape such that the shape ...
详细信息
An algorithm for fitting a conic shape to a set of given data is described. By a modified medial axis transform, described in the paper, the algorithm first determines the orientation of the shape such that the shape could be rotated to align its axis withthe X-axis. the standard nonlinear equations of the conic, both closed and open, are then fitted using a linearized least square approach. Some experimental results are provided indicating the convergence and good performance of the algorithm.< >
the performance of the various parts of the HERMES multiprocessor vision system is evaluated. HERMES is an autonomous, hierarchical, heterogenic visionprocessing system, consisting of N/sup 2//4/sup i/, 0 >
the performance of the various parts of the HERMES multiprocessor vision system is evaluated. HERMES is an autonomous, hierarchical, heterogenic visionprocessing system, consisting of N/sup 2//4/sup i/, 0 >
暂无评论