Hitachi-kokusai Electric Inc.

Shinya Ogura

Seiichi Hirai

Wataru Ito

Hirotada Ueda

Satoshi Maruta

Yuichiro Komiya

Hiroto Nagayoshi

Tomomi Takada

Shigeyuki Murata

Ryosuke Kimura

Takashi Mito

Naoyuki Shimbo

Sho Nishino

Toshiaki Hirama

A similar image search system is provided and searches for images similar to a search key image, where a number of searched images of person are classified into a number of groups based on a similarity between the respective images. A grouping display is performed in which at least one of the images that belongs to each group is displayed in a display region that is provided for each group. The grouping display is performed in such a manner that the order of the similarity between each group and the search key image can be visually recognized.

Similar image search system

An image processing device or a method of processing an image is disclosed. The method includes receiving an image, detecting a plurality of person images in the image, identifying at least a first person image from among the plurality of person images by a preferential method, identifying at least a second person image from among the plurality of person images excluding the first person image, by an exhaustive method, and extracting a first feature amount as to the first person image and a second feature amount as to the second person image during a time interval. For example, the total number of the first person images and the second person images is no more than a maximum number of person images, identified by performance corresponding to process the person images during the time interval.

Image processing device and image processing system

Disclosed is a similar facial image search system which is less susceptible to bias in search results even when there is a difference in the angle from which a face is photographed. The present invention is provided with: a means for detecting a first face from the inputted search subject image; a means for calculating and a means for recording a feature value of the detected first face; a means for detecting a second face from an inputted key image used for search; a means for calculating the face angle of the second face; a means for determining a synthesis pattern from the calculated face angle; a means for generating a synthetic face image according to the determined synthesis pattern; a means for calculating the feature value of the second face using the generated synthetic face image; a means for searching a database using a plurality of calculated face feature values as a query; and a means for integrating the plurality of the retrieved search results.

An image reproduction function for reproducing effectively reproducing image recorded in a storage device in diversified forms is provided. A moving picture obtained from one image pickup device is recorded in a plurality of systems at different frame rates. The recorded moving picture is formed of a plurality of still picture data. When reproducing the recorded moving picture, the still picture data forming a moving picture of either one of the systems are reproduced according to a time axis serving as reference for reproduction of the recorded moving picture.

Image display method, image display device, and image display program

An image storage system effectively stores image data captured by an image capture device in a plurality of image storage devices. Each of the image storage devices includes a storage unit for storing image data. An image storage device performing as a master among the image storage devices includes an image reception unit for receiving the image data captured by the image capture device. A storage unit provided in each of the image storage devices distributes the image data received by the image reception unit provided in the image storage device performing as the master to the respective image storage devices so as to be stored there.

Image storage system

Object detection method and apparatus, including an image-pickup unit picking up the image of a target region, a first recording device accumulating, as a reference image, an image signal from the image-pickup unit, a second recording device storing a table which defines surveillance information on the image signal, an object detection processing unit detecting a target object from the image signal and the reference image, a memory table storing a processing result from the processing unit, and a display unit. The image signal from the image-pickup unit is saved based on the table of the second recording device, and the processing result in the memory table is displayed on the display unit. The surveillance information in the table of the second recording device is modified based on the display of the processing result obtained by the object detection processing unit.

Object detection method and apparatus which perform object detection using set parameters

In object detection method and apparatus, a detection image related to a binary image obtained by subtracting a background image from an input image delivered from an image-pickup device is displayed on the display screen of a display unit, responsive to an operation done by an operator who determines that at least one object in the displayed detection image is a region to be excluded from object detection, the region of the object is set as a masking region automatically and thereafter, object detection is performed in respect of individual input images while performing masking in the masking region.

Object detection method using an image-pickup device with easy detection masking region setting and object detection apparatus using the method

An object is detected in comparison between an input image from an image pick-up device having a zoom mechanism and a template image stored in a manner that a first image within a view field of the image pick-up device is stored as a first as the template image, a power of the zoom mechanism to be changed is recorded, a second image to be detected is picked-up from the image pick-up device. Then, a size of either one of the template image and the second image is changed on the bases of the changed power of the zoom mechanism, and a template matching is performed between the template image and the second image to detect the object. This process makes it possible to track an object within the view field.

Growing community of inventors

Kodaira, Japan

Shinya Ogura