The effect of ambient light on a measurement taken by an imaging pixel may be reduced by employing two optical filters. The two filters may have narrow passbands that are close to each other but do not overlap. The first filter may allow ambient and active light to pass. The second filter may allow ambient light to pass but may block active light. The ambient and active light that passes through the first filter may cause electrical charge to be generated in a photodiode of the pixel. The ambient light that passes through the second filter and strikes another pixel element may control the amperage of an electrical current that depletes charge from the photodiode. For instance, the other element may be a photoresistor, the light-dependent resistance of which controls the amperage, or may be a second photodiode that generates charge that controls a transistor that controls the amperage.

There is provided an image sensor for exposing a plurality of pixel rows within a frame period using a rolling shutter. The image sensor includes a processor for calculating bright-dark distribution patterns of image frames. The processor further adjusts the frame period to be substantially identical to a predetermined period by changing a total number of exposed line times within the frame period when a difference between the bright-dark distribution patterns of two image frames is larger than a predetermined threshold.

An image sensor includes a pixel array; a logic circuit configured to convert an image signal generated from the pixel array during a first period into image data; and a memory. The image data may be written in the memory during a second period, of which at least a portion overlaps the first period. The logic circuit may write dummy data in the memory during a third period overlapping the first period and not overlapping the second period.

Techniques are described for sampled bandgap reference generation for CMOS image sensor (CIS) applications. For example, the CIS includes a pixel array, one or more pixel analog to digital converters (ADCs), and a sampled bandgap reference generator, all integrated in close proximity on a chip. The ADCs rely on stable reference levels from the bandgap reference generator for performing pixel conversions for the pixel array. Embodiments of the sampled bandgap reference generator can operate according to reference generation cycles. Each cycle can include a first portion, in which an active core dynamically stabilizes the bandgap reference level; and a second portion, in which the core is deactivated, and the bandgap reference level is output based on a sampled level obtained during the preceding first portion of the cycle. The cycle timing can be controlled to achieve sufficient dynamic stabilization of the reference levels, while mitigating photon emissions from the core.

An image signal processor and an image sensor including the same are disclosed. An image sensor includes a pixel array configured to convert received optical signals into electrical signals, a readout circuit configured to convert the electrical signals into image data and output the image data, and an image signal processor configured to perform deep learning-based image processing on the image data based on training data selected from among first training data and second training data based on a noise level of the image data.

An image signal processor and an image sensor including the same are disclosed. An image sensor includes a pixel array configured to convert received optical signals into electrical signals, a readout circuit configured to convert the electrical signals into image data and output the image data, and an image signal processor configured to perform deep learning-based image processing on the image data based on training data selected from among first training data and second training data based on a noise level of the image data.

An imaging device includes a first substrate, a second substrate, a third substrate, and a switching unit. The first substrate has a pixel including a photodiode and floating diffusion that holds the charge converted by the photodiode. The second substrate has a pixel circuit that reads out a pixel signal based on the charge held in the floating diffusion in the pixel, and is stacked on the first substrate. The third substrate has a processing circuit that detects a pixel signal read out by the pixel circuit, and is stacked on the second substrate. The switching unit is provided to enable electrical connection between the floating diffusion and a floating diffusion of another pixel in the first substrate, and is provided on the second substrate. As a result, by switching the capacitance of the floating diffusion of the pixel using floating diffusion of another pixel, it is possible to switch the charge-voltage conversion efficiency levels.

A solid-state imaging apparatus includes: an imaging section that acquires image data; and a control section that causes DNN processing on the image data and readout processing of the image data to be executed in parallel and causes noise reduction processing to be executed on the image data that is read out when the DNN processing on the image data and the readout processing of the image data are being executed in parallel.

An image-capturing device includes: an image sensor that includes an image capturing area where an image of a subject is captured; a setting unit that sets image capturing conditions to be applied to the image-capturing area; a selection unit that selects pixels to be used for interpolation from pixels present in the image-capturing area; and a generation unit that generates an image of the subject captured in the image-capturing area with signals generated through interpolation executed by using signals output from the pixels selected by the selection unit, wherein: the selection unit makes a change in selection of at least some of the pixels to be selected depending upon the image-capturing conditions set by the setting unit.

Systems and methods are disclosed for image signal processing. For example, methods may include receiving a current image of a sequence of images from an image sensor; combining the current image with a recirculated image to obtain a noise reduced image, where the recirculated image is based on one or more previous images of the sequence of images from the image sensor; determining a noise map for the noise reduced image, where the noise map is determined based on estimates of noise levels for pixels in the current image, a noise map for the recirculated image, and a set of mixing weights; recirculating the noise map with the noise reduced image to combine the noise reduced image with a next image of the sequence of images from the image sensor; and storing, displaying, or transmitting an output image that is based on the noise reduced image.