Samsung Patent | Depth sensor, image capture method, and image processing system using depth sensor

Patent: Depth sensor, image capture method, and image processing system using depth sensor

Publication Number: 10171790

Publication Date: 2019-01-01

Applicants: Samsung

Abstract

An image capture method performed by a depth sensor includes; emitting a first source signal having a first amplitude towards a scene, and thereafter emitting a second source signal having a second amplitude different from the first amplitude towards the scene, capturing a first image in response to the first source signal and capturing a second image in response to the second source signal, and interpolating the first and second images to generate a final image.

Background

The inventive concept relates generally to sensors and methods of operating sensors. More particularly, the inventive concept relates to depth information calculations, depth sensors that use a time of flight (TOF) principle, image capture methods using a depth sensor, and image processing systems including a depth sensor.

A “sensor” is a device that detects the state of an object and transforms a detection result into a corresponding electrical signal. Certain sensors are also capable of communicating the corresponding electrical signal to a variety of external circuits.

Sensors include light sensors, temperature sensors, pressure sensors, magnetic sensors, and so-called depth (or distance) sensors. The term “depth sensor” is used to denote a class of sensors that detects–as one type of state–the location (or relative location) of the object. Sensors may be used to detect the state of an object in relation to a particular range of electromagnetic signals, such as microwave signals, visible light signals, infrared signals, ultrasonic signals, etc. In certain applications, a sensor will be associated with a “source” that transmits (or emits) a “source signal” or ranging signal towards an object. The object may then reflect a portion of the source signal, and the reflected portion of the source signal is detected by a depth sensor.

In this manner, a depth sensor may measure a depth (or range. or distance) between the object and the sensor using a TOF measuring method. That is, the depth sensor may be used to measures a delay time between the transmission (or emission) of the source signal and return of the reflected portion of the source signal to the sensor. In this context, an area surrounding the object, an area effectively receiving transmission of the source signal, and/or an area effectively detected by the sensor may be termed a “scene”.

When a depth sensor is disposed relatively far from a scene, the level of the source signal reaching the scene and/or the level of the reflected portion of the source signal returned to the depth sensor may be relatively low. A low source signal level adversely affects the signal-to-noise ratio for the reflected portion of the source signal. That is, the depth sensor will received a relatively large quantity of noise along with the relatively weak reflected portion of the source signal.

In contrast, when the depth sensor is relatively close to a scene, the level of the reflected portion of the source signal will be relatively high, thereby providing a good signal-to-noise ratio. Unfortunately, many scenes contain some objects that are far from the depth sensor and other objects that are near the depth sensor. And it is often difficult to capture images from a complex assortment of signals reflected from both near and far objects or scene portions. Accordingly, there is a demand for methods that increase the overall accuracy of depth information provided by depth sensors, particularly as such depth information relates to scenes having multiple objects separated near and far from the depth sensor.

Summary

According to an aspect of the inventive concept, there is provided an image capture method performed by a depth sensor, the method comprising emitting source signals having different amplitudes to a scene sequentially; and capturing images according to source signals sequentially reflected by the scene.

The image capture method may further comprise generating a single image by interpolating the images.

An source signal having a low amplitude from among the source signals is used to capture a point on the scene that is close to the depth sensor. An source signal having a high amplitude from among the source signals is used to capture a point on the scene that is far from the depth sensor.

According to another aspect of the inventive concept, there is provided a depth sensor comprising a light source which emits source signals having different amplitudes to a scene sequentially; a light source driver which drives the light source sequentially so that the source signals have different amplitudes; and a depth pixel which detects pixel signals having different pixel values according to source signals sequentially reflected by the scene.

The depth sensor may further comprise an image signal processor which generates images by using the pixel signals having the different pixel values.

The image signal processor may generate a single image by interpolating the images. The light source may be an infrared diode or a laser diode.

According to another aspect of the inventive concept, there is provided an image processing system comprising the depth sensor; and a processor which processes pixel signals output by the depth sensor.

The image processing system may be a portable device. The image processing system may be a smart TV, a handheld game console, or a security camera.

发表评论

电子邮件地址不会被公开。 必填项已用*标注