Abstract: A method and system of evaluating a valid analysis region of a specific scene, wherein the method and system performs image analyses on continuous images/frames of a specific scene to obtain detectable object or event information therein, so as to generate a closed valid analysis region to reduce the overall data and loading of image analyses during actual monitoring, processing and analyzing of the specific scene.

Abstract: The present invention discloses a method of evaluating image quality taken in an actual scene, the method comprising: providing a test chart placed in an actual scene, said test chart comprising a plurality of positioning patterns and a plurality of quality test patterns; capturing an image of the test chart in the actual scene; searching the plurality of positioning patterns in the image of the test chart to locate the orientation of the image of the test chart; and identifying the plurality of quality test patterns in the located image of the test chart and evaluating the image quality of the plurality of quality test patterns.

Abstract: The present invention discloses a method of evaluating image quality taken in an actual scene, the method comprising: providing a test chart placed in an actual scene, said test chart comprising a plurality of positioning patterns and a plurality of quality test patterns; capturing an image of the test chart in the actual scene; searching the plurality of positioning patterns in the image of the test chart to locate the orientation of the image of the test chart; and identifying the plurality of quality test patterns in the located image of the test chart and evaluating the image quality of the plurality of quality test patterns.

Abstract: A method for face recognition based on online learning by calculating a similarity between each of the face images captured in a specific environment and a target face image so as to form a distribution of the similarities of the face images with respect to the target face image, wherein a similarity threshold with respect to the first target face image is determined according to a predefined rule and the distribution of the similarities, wherein the similarity threshold is used for subsequent selection of a face image captured in the specific environment and having a similarity greater than said similarity threshold with respect to the target face image.

Abstract: The invention discloses a system and method for estimating a three-dimensional (3D) packaging size of an object. The method has a self calibration function to correct the scale in the measurement environment so as to reduce errors caused by human operations. The system comprises an image capture unit for capturing images of the measurement environment and objects; a scale correction unit for performing the calibration using a reference object of a known size; and an object size estimation unit for detecting the position of an object to be estimated in the image data received from the image capture unit and estimating the 3D size of the object according to the calibrated scale.

Abstract: A method and system for generating a video synopsis from compressed video images is disclosed, wherein the compressed video images are generated by dividing each image of the original video into a plurality of macroblocks, wherein each of the macroblocks has the same macroblock size. By analyzing and utilizing characteristics of video compression information to determine the importance of each of the macroblocks, the distribution of the important macroblocks in time and space can be obtained, and a video synopsis of a specified length can be synthesized by arranging the important macroblocks in a particular temporal and spatial order.

Abstract: Disclosed are an automatic media editing apparatus, an editing method, a media broadcasting method and a broadcasting system thereof. The media editing apparatus edits or labels the media clips in accordance with specified features according to the analysis of the received multimedia. A new multimedia clip can be created after reproducing the media clips. The sources of the multimedia may be the contents which are side-recorded by a plurality of area servers located at multiple areas. Followed on, the edited multimedia as required can be broadcasted to clients. The automatic media editing technology may be applied to a media broadcasting system across multiple areas. The system may include a main server and a plurality of area servers. The area server may integrate and clip the content in accordance with the specified features.

Abstract: The invention discloses a system and method for estimating a three-dimensional (3D) packaging size of an object. The method has a self calibration function to correct the scale in the measurement environment so as to reduce errors caused by human operations. The system comprises an image capture unit for capturing images of the measurement environment and objects; a scale correction unit for performing the calibration using a reference object of a known size; and an object size estimation unit for detecting the position of an object to be estimated in the image data received from the image capture unit and estimating the 3D size of the object according to the calibrated scale.

Abstract: Described herein are a system and a method for abnormal behavior detection using automatic classification of multiple features. Features from various sources, including those extracted from camera input through digital image analysis, are used as input to machine learning algorithms. These algorithms group the features and produce models of normal and abnormal behaviors. Outlying behaviors, such as those identified by their lower frequency, are deemed abnormal. Human supervision may optionally be employed to ensure the accuracy of the models. Once created, these models can be used to automatically classify features as normal or abnormal. This invention is suitable for use in the automatic detection of abnormal traffic behavior such as running of red lights, driving in the wrong lane, or driving against traffic regulations.

Abstract: The invention discloses a correction method and system for object linking in video sequences that are captured by a multiple camera surveillance video system. The invention allows a user to select a specific object to track and for trace correction through an interactive user platform. The interactive user platform shows a video sequence captured by the user-selected camera at the selected time; lists of previous and post video sequences related to the user-selected object or video sequences, and lists of object linking results before and after the selected time. The user refers to the linking results, shown in lists of object linking results before and after the selected time to select one of objects for correction in selected frame of selected video sequence. The selection directs the proposed system to correct the trace of the specific object to track. Henceforth, the correct linking results can be automatically generated.

Abstract: The invention discloses a correction method and system for object linking in video sequences that are captured by a multiple camera surveillance video system. The invention allows a user to select a specific object to track and for trace correction through an interactive user platform. The interactive user platform shows a video sequence captured by the user-selected camera at the selected time; lists of previous and post video sequences related to the user-selected object or video sequences, and lists of object linking results before and after the selected time. The user refers to the linking results, shown in lists of object linking results before and after the selected time to select one of objects for correction in selected frame of selected video sequence. The selection directs the proposed system to correct the trace of the specific object to track. Henceforth, the correct linking results can be automatically generated.

Abstract: Disclosed are an automatic media editing apparatus, an editing method, a media broadcasting method and a broadcasting system thereof. The media editing apparatus edits or labels the media clips in accordance with specified features according to the analysis of the received multimedia. A new multimedia clip can be created after reproducing the media clips. The sources of the multimedia may be the contents which are side-recorded by a plurality of area servers located at multiple areas. Followed on, the edited multimedia as required can be broadcasted to clients. The automatic media editing technology may be applied to a media broadcasting system across multiple areas. The system may include a main server and a plurality of area servers. The area server may integrate and clip the content in accordance with the specified features.

Abstract: Disclosed herein are a system and method for the automatic detection of traffic and parking violations. Camera input is digitally analyzed for vehicle type and location. This information is then processed against local traffic and parking regulations to detect violations. Detectable driving offenses include, but are not limited to: no scooters, buses only, and scooters only lane violations. Detectable parking offenses include, but are not limited to: parking or loitering in bus stops, parking next to fire hydrants, and parking in no-parking zones. Camera input, detected vehicle information, and violations can be stored for later search and retrieval. The system may be configured to signal the authorities or other automated analysis systems about specific violations. When coupled with automatic license plate recognition, vehicles may be automatically matched against a registration database and reported or ticketed.

Abstract: Described herein are a system and a method for abnormal behavior detection using automatic classification of multiple features. Features from various sources, including those extracted from camera input through digital image analysis, are used as input to machine learning algorithms. These algorithms group the features and produce models of normal and abnormal behaviors. Outlying behaviors, such as those identified by their lower frequency, are deemed abnormal. Human supervision may optionally be employed to ensure the accuracy of the models. Once created, these models can be used to automatically classify features as normal or abnormal. This invention is suitable for use in the automatic detection of abnormal traffic behavior such as running of red lights, driving in the wrong lane, or driving against traffic regulations.