Provided is an image-capturing device including: an image-capturing unit that acquires an image of a subject; a subject-extracting unit that separates and extracts a main subject and a background from the acquired image; a distance-information acquiring unit that calculates the distance between the image-capturing unit and the extracted main subject and the distance between the image-capturing unit and the extracted background; a photographing-condition acquiring unit that acquires acquisition-time photographing conditions for an acquisition time of the image and an adjustable photographing condition that can be adjusted with respect to the acquisition-time photographing conditions; a photographing-condition selecting unit that selects a photographing condition that is worth adjusting on the basis of the acquired acquisition-time photographing conditions and adjustable photographing condition and the acquired distances; and an image-processing unit that generates, by performing image processing on the acquired image, an image that will be acquired when the selected photographing condition is set.

Provided is an image-capturing device including: an image-capturing unit that acquires an image of a subject; a subject-extracting unit that separates and extracts a main subject and a background from the acquired image; a distance-information acquiring unit that calculates the distance between the image-capturing unit and the extracted main subject and the distance between the image-capturing unit and the extracted background; a photographing-condition acquiring unit that acquires acquisition-time photographing conditions for an acquisition time of the image and an adjustable photographing condition that can be adjusted with respect to the acquisition-time photographing conditions; a photographing-condition selecting unit that selects a photographing condition that is worth adjusting on the basis of the acquired acquisition-time photographing conditions and adjustable photographing condition and the acquired distances; and an image-processing unit that generates, by performing image processing on the acquired image, an image that will be acquired when the selected photographing condition is set.

A depth estimation method for a monocular image based on a multi-scale CNN and a continuous CRF is disclosed in this invention. A CRF module is adopted to calculate a unary potential energy according to the output depth map of a DCNN, and the pairwise sparse potential energy according to input RGB images. MAP (maximum a posteriori estimation) algorithm is used to infer the optimized depth map at last. The present invention integrates optimization theories of the multi-scale CNN with that of the continuous CRF. High accuracy and a clear contour are both achieved in the estimated depth map; the depth estimated by the present invention has a high resolution and detailed contour information can be kept for all objects in the scene, which provides better visual effects.

A depth estimation method for a monocular image based on a multi-scale CNN and a continuous CRF is disclosed in this invention. A CRF module is adopted to calculate a unary potential energy according to the output depth map of a DCNN, and the pairwise sparse potential energy according to input RGB images. MAP (maximum a posteriori estimation) algorithm is used to infer the optimized depth map at last. The present invention integrates optimization theories of the multi-scale CNN with that of the continuous CRF. High accuracy and a clear contour are both achieved in the estimated depth map; the depth estimated by the present invention has a high resolution and detailed contour information can be kept for all objects in the scene, which provides better visual effects.

The present disclosure provides a detection system, which includes an image sensor, a lens device, and a processor. The image sensor is configured to take a first picture of a foreground object and a background object. The lens device is attached to the image sensor and configured to allow the foreground object to form a clear image on the first picture and the background object to form a blurred image on the first picture. The processor is configured to determine the image of the foreground object by analyzing the sharpness of the images of the first pictures.

An imaging device includes a focus driving unit for driving a focus lens; a tilting driving unit for tilting at least one of an image sensor and an imaging optical system; a detection unit for detecting a subject in an image; a setting unit for setting a plurality of focus regions in the image based on positional information of the subject detected in the image; a determination unit for determining respective driving amounts for the focus driving unit and for the tilting driving unit based on the positional information and focal information of the plurality of focus regions so that the plurality of focus regions is in focus; and a control unit for controlling the focus driving unit and the tilting driving unit based on the determined respective driving amounts.

The present disclosure provides a detection system, which includes an image sensor, a lens device, and a processor. The image sensor is configured to take a first picture of a foreground object and a background object. The lens device is attached to the image sensor and configured to allow the foreground object to form a clear image on the first picture and the background object to form a blurred image on the first picture. The processor is configured to determine the image of the foreground object by analyzing the sharpness of the images of the first pictures.

Systems and methods described herein can be used to improve camera modules (e.g., camera components), particularly when the camera lens depth of focus is very small. Improvement of a camera with a small depth of focus is particularly important in various applications, such as in cameras used in autonomous navigation (e.g., advanced driver assistance systems (ADAS) and autonomous vehicle (AV) systems).

Systems and methods described herein can be used to improve camera modules (e.g., camera components), particularly when the camera lens depth of focus is very small. Improvement of a camera with a small depth of focus is particularly important in various applications, such as in cameras used in autonomous navigation (e.g., advanced driver assistance systems (ADAS) and autonomous vehicle (AV) systems).

Aspects of the present disclosure are directed to a method for focusing a camera. In one embodiment, the method includes: dividing the field of view of the camera in to at least two segments: assigning, in each case, at least one operating element or at least one position of an operating element to the at least two segments; recognizing and tracking at least one object in at least two segments; automatically assigning the recognized at least two objects to the respective operating element or position of the operating element depending on which segment the objects are assigned to; and focusing the camera on the object assigned to the operating element or the position of the operating element in response to the operating element being actuated or the operating element being brought into the corresponding position.