Canon Kabushiki Kaisha

Ka Ming Leung

Getian Ye

Sammy Chan

Mark Ronald Tainsh

Geoffrey Richard Taylor

Ian Robert Boreham

A method of generating a training sample for matching unlabelled objects in a sequence of images. A first representation of unlabelled objects is generated from images of a first and second set of images. A second representation of the unlabelled objects is generated using an unsupervised method. An anchor image in the first set is selected. A first set of candidate images in the second set that are close to the anchor image in both the first and second representations, is determined. A second set of candidate images in the second set that are distant from the anchor image in either the first or the second representations, is determined. A match candidate image is selected from the first set or the second set of candidate images. The training sample is generated from at least the anchor image and the match candidate image.

Method, system and apparatus for generating training samples for matching objects in a sequence of images

A method of comparing objects in images. A dictionary determined from a plurality of feature vectors formed from a test image and codes formed by applying the dictionary to the feature vectors, is received. The dictionary is based on a modified manifold obtained by determining correspondences for codes using pairwise similarities between codes. Comparison codes are determined for the objects in the images by applying the dictionary to feature vectors of the objects in the images. The objects in the images are compared based on the comparison codes of the objects.

Method, system and apparatus for comparing objects in images

A method of selecting at least one video frame of a video sequence comprising a plurality of video frames. The method determines a time for analysis based on length of the video sequence and processing capability of a running device. A first sampling pattern is based on the determined time for analysis. A first set of frames in the video sequence is sampled using a first sampling pattern. The first set is sampled infrequently throughout the video sequence in accordance with the first sampling pattern. A candidate frame is determined from the sampled frames based on image quality. A second set of the frames comprising one or more of the frames in a narrow range of the video sequence near the determined candidate frame, is determined in accordance with a second sampling pattern. At least one of the video frames is selected from the sampled frames based on image quality.

Method, system and apparatus for selecting a video frame

A method of determining similarity of objects in images. Feature vectors are determined for objects in images captured by cameras operating in a training domain. Feature vectors are determined for the objects in images captured by cameras operating in a target domain, the cameras of the target domain operating with different environmental factors to the cameras of the training domain. A mapping is determined for a difference in the feature vectors of the training domain and the target domain. The difference in the feature vectors of the training domain and the target domain is converted to a matching space by applying the determined mapping to the feature vectors of the training domain and the target domain. A classifier is determined using data associated with the feature vectors of the training domain in the matching space. A similarity of the objects in the images captured in the target domain is determined using a difference of feature vectors for the objects by applying the classifier to the feature vectors of the objects in the matching space.

Method and apparatus for determining similarity of objects in images

A method of selecting at least one video frame of a video sequence. A plurality of faces is detected in at least one video frame of the video sequence. An orientation of the detected faces is tracked over a series of subsequent video frames to determine whether a first detected face is turning towards a second detected face. The method then determines, using the tracked orientation of the detected faces, a portion of the video sequence in which the first and second detected faces are oriented towards each other for at least a predetermined number of frames defining a gaze fixation of the detected faces. At least one video frame is selected from the determined portion of the video sequence, the selected video frame capturing the gaze fixation of the detected faces.

Methods, systems, and computer readable media are described for controlling a camera to perform a selected task from a set of tasks is provided. The method comprises determining a viewing condition of the camera to perform each task from the set of tasks, and determining a posterior probability of task success for each task from the set of tasks based on the determined viewing conditions and a prior probability of task success for each task. The method also includes determining a change in rate of information gain for task success for each task from the set of tasks based on the posterior probability, selecting the task to be performed based on the change in rate of information gain, and controlling the camera to perform the selected task.

Methods and systems for controlling a camera to perform a task

A method of selecting a frame from a plurality of video frames captured by a camera (). The method determines features to which map points in a three dimensional space are projected. A histogram of the determined features for a plurality of regions in the frame is created. One of the regions may be determined as being an unmapped region based on the created histogram. The frame is selected based on the size of the unmapped region.

Growing community of inventors

Marsfield, Australia

Ka Ming Leung