Abstract: A temperature distribution learning apparatus learns a temperature distribution including a temperature at at least a first point in air of a target space. The temperature distribution learning apparatus includes a learning model. The learning model learns the temperature distribution, which is an objective variable, and a thermal image, which is an explanatory variable and is related to the target space, in association with each other. The learning model learns based training data including the temperature distribution and the thermal image.

Abstract: The present invention discloses a capsule endoscope image display controller (26) including: an image-to-image similarity calculating unit (36) that calculates, for each image included in an image sequence captured by a capsule endoscope which moves within the digestive organs, a similarity between the image and its temporally consecutive image; an amount-of-movement calculating unit (47) that calculates, for each image included in the image sequence, an amount of movement of a feature area included in the image; a video state classifying unit (41) that classifies, for each image included in the image sequence, a video state of the image into one of the following states, based on the video state, the similarity, and the amount of movement of the image: (a) “stationary state” indicating that the capsule endoscope is stationary, (b) “digestive organs deformation state” indicating that the digestive organs are deformed, and (c) “capsule moving state” indicating that the capsule endoscope is moving, based on the

Abstract: Meta data about content is converted into characteristic values. Each characteristic value is associated with one or more content segments of content. The content can be audio, video, or other data or combinations of data. Importance levels for content segments are determined from content scores. Content scores are generally an indication of how one person would rate a particular characteristic value. The content scores may be estimated by using previously determined scores of similar content segments. The similarity is preferably determined by a particular metric. A user may also supply his or her own content scores for content segments. The user profile content scores can be combined with the determined content scores or completely supplant these scores. Based on the importance levels for content scores for content segments, content segments may be packaged into a content digest that a user may view.

Abstract: An endoscope that is free from a dead area and capable of preventing the physician from overlooking any nidus is an endoscope for taking the inside of digestive organs, and the endoscope is provided with an omnidirectional camera (32), a light (34), a forceps (36) and a rinse water injection port (38) at the tip (24). The omnidirectional camera (32) is a device for taking the inside of digestive organs, and is able to take 360-degree images of its surroundings. A probe-type endoscope (20) is provided with a receiver (26) composed of orthogonal coils, and the receiver (26) is used for estimating the position and attitude of the probe-type endoscope (20). An image taken by the omnidirectional camera (32) is presented on a display unit (28) of an image processing device (22) connected to the probe-type endoscope (20).

Abstract: The present invention provides methods for extracting an image region segment from a video image, a system for extracting an image region segment, a program for extracting an image region segment, a method for distributing an extracted video image, and a method for generating content. Motion compensation data is calculated using sequential video images captured by a video camera, and the sequential video images are employed to prepare an estimation of the video camera operation. Based on the estimated camera operation, the center position (xc, yc) of an area that is aimed at in a video image is estimated under predetermined conditions (rules), an image feature is designated in the vicinity of the center position of the target area, and an image segment is specified that includes this image feature.

Abstract: To implement more appropriate QoS control by compressing contents data distributed via a network taking into consideration meaning of contents and preferences of users. A network system comprises a content provider that provides content and client that acquires the content. The content provider attaches information on communication priorities to the content based on the content's importance which is determined from the meaning of the content's and sends the content to the client using a protocol which controls quality of service based on the priority and the client receives the content transmitted by the content provider.

Abstract: The present invention provides methods for extracting an image region segment from a video image, a system for extracting an image region segment, a program for extracting an image region segment, a method for distributing an extracted video image, and a method for generating content. Motion compensation data is calculated using sequential video images captured by a video camera, and the sequential video images are employed to prepare an estimation of the video camera operation. Based on the estimated camera operation, the center position (xc, yc) of an area that is aimed at in a video image is estimated under predetermined conditions (rules), an image feature is designated in the vicinity of the center position of the target area, and an image segment is specified that includes this image feature.

Abstract: Methods and apparatus for effectively identifying the occlusion of objects, such as persons, having a high degree of freedom. In an example embodiment, after initialization, an image is input, and an image region is extracted from image data. The distance is employed that is obtained when the shape of a two-dimensional histogram in the color space is transformed into the feature space. A graph is formed by using the regions between the frames. A confidence factor is provided and image features are provided as weights to the edges that connect the nodes. Processing is performed and the confidence factor is examined. A connection judged less possible to be a path is removed. When there is only one available connection for the occlusion point, this connection is selected.

Abstract: To implement more appropriate QoS control by compressing contents data distributed via a network taking into consideration meaning of contents and preferences of users. A network system comprises a content provider that provides content and client that acquires the content. The content provider attaches information on communication priorities to the content based on the content's importance which is determined from the meaning of the content's and sends the content to the client using a protocol which controls quality of service based on the priority and the client receives the content transmitted by the content provider.

Abstract: The present invention discloses a capsule endoscope image display controller (26) including: an image-to-image similarity calculating unit (36) that calculates, for each image included in an image sequence captured by a capsule endoscope which moves within the digestive organs, a similarity between the image and its temporally consecutive image; an amount-of-movement calculating unit (47) that calculates, for each image included in the image sequence, an amount of movement of a feature area included in the image; a video state classifying unit (41) that classifies, for each image included in the image sequence, a video state of the image into one of the following states, based on the video state, the similarity, and the amount of movement of the image: (a) “stationary state” indicating that the capsule endoscope is stationary, (b) “digestive organs deformation state” indicating that the digestive organs are deformed, and (c) “capsule moving state” indicating that the capsule endoscope is moving, based on the

Abstract: The present invention provides methods for extracting an image region segment from a video image, a system for extracting an image region segment, a program for extracting an image region segment, a method for distributing an extracted video image, and a method for generating content. Motion compensation data is calculated using sequential video images captured by a video camera, and the sequential video images are employed to prepare an estimation of the video camera operation. Based on the estimated camera operation, the center position (xc, yc) of an area that is aimed at in a video image is estimated under predetermined conditions (rules), an image feature is designated in the vicinity of the center position of the target area, and an image segment is specified that includes this image feature.

Abstract: Methods and apparatus for effectively identifying the occlusion of objects, such as persons, having a high degree of freedom. In an example embodiment, after initialization, an image is input, and an image region is extracted from image data. The distance is employed that is obtained when the shape of a two-dimensional histogram in the color space is transformed into the feature space. A graph is formed by using the regions between the frames. A confidence factor is provided and image features are provided as weights to the edges that connect the nodes. Processing is performed and the confidence factor is examined. A connection judged less possible to be a path is removed. When there is only one available connection for the occlusion point, this connection is selected.

Abstract: An endoscope that is free from a dead area and capable of preventing the physician from overlooking any nidus is an endoscope for taking the inside of digestive organs, and the endoscope is provided with an omnidirectional camera (32), a light (34), a forceps (36) and a rinse water injection port (38) at the tip (24). The omnidirectional camera (32) is a device for taking the inside of digestive organs, and is able to take 360-degree images of its surroundings. A probe-type endoscope (20) is provided with a receiver (26) composed of orthogonal coils, and the receiver (26) is used for estimating the position and attitude of the probe-type endoscope (20). An image taken by the omnidirectional camera (32) is presented on a display unit (28) of an image processing device (22) connected to the probe-type endoscope (20).

Abstract: The object of the invention is to extract from a moving picture and display at least a portion of team play, especially in a field sport, that is considered to be important for detailed analysis, so that the teamwork involved is easily discerned.

Abstract: An object of the present invention is to provide a description method for efficiently representing contents of motion picture with a small data volume. The organization of the present invention (1) represents a trajectory of how each object has moved over time by using reference plane representing position information of each object, (2) sets a description unit based on a type of action of each object by using changes in shape of each object, (3) has actions of each object represented as each behavioral section, and (4) comprises a description facility capable of reading and interpreting definition of an object dependent on video contents, definition of classes of actions, and definition of interpretation of a scene by interaction of plural objects.

Abstract: The invention provide methods and apparatus for effectively identifying the occlusion of objects, such as persons, having a high degree of freedom. In an example embodiment, after initialization, an image is input, and an image region is extracted from image data. The distance is employed that is obtained when the shape of a two-dimensional histogram in the color space is transformed into the feature space. A graph is formed by using, the regions between the frames. A confidence factor is provided and image features are provided as weights to the edges that connect the nodes. Processing is performed, and the confidence factor is examined. A connection judged less possible to be a path is removed. When there is only one available connection for the occlusion point, this connection is selected.

Abstract: An index generator that generates an index, which is data description contents, such as video contents, comprises: an index description device, for defining in advance basic index information concerning an index; a video display device for the input, the display or the output of contents to which an index is to be added; a triggering action input device, for accepting a triggering action in the contents that is displayed or output; and an index determination device, for generating index data based on the basic index information, which is defined by the index description device, and triggering action input history information, which is entered by the triggering action input device.

Abstract: A method and an apparatus for using the trajectory of an object to access video contents, for example, to specify and display a specific video image scene. Such a video contents access method comprises the steps of: extracting objects from video contents; displaying movements of the objects as trajectories on a specific projection screen; specifying locations on the trajectories; and accessing a desired scene of the video contents. An apparatus is so designed that it performs the above method.

Abstract: To implement more appropriate QoS control by compressing contents data distributed via a network taking into consideration meaning of contents and preferences of users.

Abstract: The invention provides a method for classifying the motion of an object such as a human being in a moving picture. A template is prepared in advance, which includes the Gabor wavelet expansion coefficients of an object image in a plurality of frames of a video image sequence representing each of a plurality of different reference motions of an object in a moving picture. Then, processing is performed to obtain the Gabor wavelet expansion coefficients of an object image in a plurality of frames of a video image sequence representing an unknown motion of the object. Matching factors are calculated based on the expansion coefficients for the unknown motion and the expansion coefficients for the reference motions in the template, and finally the unknown motion is classified based on the matching factors.