NOCTURN Digital Extreme Low-Light CMOS Camera – Surveillance Imaging for 24/7 Lighting Conditions.
The sensor has the possibility to achieve a high optical dynamic range by using a multiple slope feature. The row selection logic supports this multiple slope operation. This feature will partially reset those pixels which reach a programmable voltage, while leaving the other pixels untouched. This can be done two times within one exposure time to achieve a maximum of three exposure slopes.
In following figure, the red lines represent a pixel in which a large amount of electrons are collected. The blue line represents a pixel with fewer electrons being collected. As shown in the figure, the bright pixel is held to a programmable voltage for a programmable time during the exposure time. This happens two times to make sure that at the end of the integration time the pixel is not saturated. The darker pixel is not influenced by this multiple slope and will have a normal response. The Vlow2/1/0 voltages and different exposure times are programmable using the on-chip registers. Using this feature, a response as detailed in next figure can be achieved. The placement of the knee points in X is controlled by the Vlow2/1/0 programming, while the slope of the segments is controlled by the programmed exposure times.
It should be noted that the knee points will have some non-uniformity on their location due to threshold voltage variations of the pixel transistors and photodiode implants.
The row logic controls the readout and integration timing of each row. The row selection logic consists of the address decoder (for start and stop addresses of the window of interest), four shift registers and a number of gates to apply the reset, select and transfer pulses. One shift register determines which row is selected for readout (a single 1 is shifted), while the other three shift register are used to define the integrating lines (consecutive 1’s are shifted) and integration conditions (multiple slope).
The address decoder is used to load pointers into the integration time shift register at the start address. The shift registers shift its value into the next row at each clk_Y pulse. After the required integration time, the readout of the window is started. This is done by loading a pointer into the read shift register. This pointer is also shifted to the next row at each clk_Y pulse. The shifting stops at the Y stop address of the window (determined by the start address ADDR_Y[10:0] and the size of the Y window SIZE_Y[10:0]). These are uploaded through the SPI interface. In this way, an electronic shutter can be realized that allows full control over transfer lines and smooth transitions of frames with varying integration time.
The rolling shutter mode in the LYNX CMOS or the KAMELEON CMOS image sensors works in the way that the photodiodes (pixels) do not collect light at the same time. All pixels in one row of the imager collect light during exactly the same period of time, but the pixel rows are reset and exposed one row after another. At the end of the exposure, the lines are read out sequentially.
Rolling shutter sensors offer a reduce sensor noise and higher pixel density compared to global shutter CMOS sensors.
The integration and readout (and therefore the rolling shutter operation) is controlled by the row logic.
The row logic has four shift registers, each containing one or more row pointers:
• The ‘read’ shift register has one pointer pointing at the row that is currently being sampled into the columns.
• The three other shift registers control the integration (each shift register controls the integration time of one of the three slopes in the response curve). The ‘inte’ shift registers point to every row that is integrating light.
The shift registers are controlled by a common clock (CLK). Each has its own sync pulse (SYNC *) to control the position of the pointer. The following table shows a timing example of the rolling shutter operation with 3 slope integration.
• The top of the figure shows the CLK and SYNC * pulses. In this example the frame size is 5 rows (1 overhead timing slot is needed every frame). The integration times of the three slopes are 4, 2 and 1.
• The bottom part of the figure shows how this timing translates to the behavior of the rows. The timing presented on the SYNC * signals is applied to the first row of the frame at the rising edge of CLK. At every edge of CLK, the integration and readout status is shifted along the image.
The sensor is integrating light between the moment the reset stops and the moment the row is sampled.
When image capture on a rolling shutter camera is triggered by an external source, achievable frame rate is half the rate achievable in free-running mode, regardless of the rate that is specified.