nqm.irimager package

Module contents

Optris PI and XI imager IR camera controller

We use the IRImagerDirect SDK (see http://documentation.evocortex.com/libirimager2/html/index.html) to control these cameras.

class nqm.irimager.IRImager

Bases: pybind11_object

IRImager object - interfaces with a camera.

get_frame(self: nqm.irimager.IRImager) → tuple[numpy.ndarray[numpy.uint16[m, n]], datetime.datetime]

Return a frame.

If the shutter is down (normally done automatically by the thermal camera for calibration), this function will wait until the shutter is back up, before returning (usually around ~1s).

Throws:

RuntimeError if a frame cannot be loaded, e.g. if the camera isn’t streaming.

Returns:

1. A 2-D matrix containing the image. This must be adjusted by get_temp_range_decimal() to get the actual temperature in degrees Celcius, offset from -100 ℃. 2. The approximate time the image was taken.

Return type:

A tuple containing

get_frame_monotonic(self: nqm.irimager.IRImager) → tuple[numpy.ndarray[numpy.uint16[m, n]], datetime.timedelta]

Return a frame, with a monotonic/steady_clock timestamp.

Similar to get_frame(), except returns a monotonic timepoint that the IRImagerDirectSDK returns, which is more accurate.

Please be aware that the epoch of the monotonic timepoint is undefined, and may be the time since last boot or the time since the program started.

get_library_version(self: nqm.irimager.IRImager) → str

Get the version of the libirimager library.

Returns:

the version of the libirmager library, or “MOCKED” if the library has been mocked.

get_temp_range_decimal(self: nqm.irimager.IRImager) → int

The number of decimal places in the thermal data

For example, if get_frame() returns 19000, you can divide this number by 10 to the power of the result of get_temp_range_decimal(), then subtract 100, to get the actual temperature in degrees Celcius.

start_streaming(self: nqm.irimager.IRImager) → None

Start video grabbing

Prefer using with irimager: … to automatically start/stop streaming on errors.

Throws:

RuntimeError if streaming cannot be started, e.g. if the camera is not connected.

stop_streaming(self: nqm.irimager.IRImager) → None

Stop video grabbing

class nqm.irimager.IRImagerMock

Bases: IRImager

Mocked version of IRImager.

This class can be used to return dummy data when there isn’t a camera connected (e.g. for testing).

get_frame(self: nqm.irimager.IRImagerMock) → tuple[numpy.ndarray[numpy.uint16[m, n]], datetime.datetime]

Return a frame.

If the shutter is down (normally done automatically by the thermal camera for calibration), this function will wait until the shutter is back up, before returning (usually around ~1s).

Throws:

RuntimeError if a frame cannot be loaded, e.g. if the camera isn’t streaming.

Returns:

1. A 2-D matrix containing the image. This must be adjusted by get_temp_range_decimal() to get the actual temperature in degrees Celcius, offset from -100 ℃. 2. The approximate time the image was taken.

Return type:

A tuple containing

get_frame_monotonic(self: nqm.irimager.IRImagerMock) → tuple[numpy.ndarray[numpy.uint16[m, n]], datetime.timedelta]

Return a frame, with a monotonic/steady_clock timestamp.

Similar to get_frame(), except returns a monotonic timepoint that the IRImagerDirectSDK returns, which is more accurate.

Please be aware that the epoch of the monotonic timepoint is undefined, and may be the time since last boot or the time since the program started.

get_temp_range_decimal(self: nqm.irimager.IRImagerMock) → int

The number of decimal places in the thermal data

For example, if get_frame() returns 19000, you can divide this number by 10 to the power of the result of get_temp_range_decimal(), then subtract 100, to get the actual temperature in degrees Celcius.

start_streaming(self: nqm.irimager.IRImagerMock) → None

Start video grabbing

Prefer using with irimager: … to automatically start/stop streaming on errors.

Throws:

RuntimeError if streaming cannot be started, e.g. if the camera is not connected.

stop_streaming(self: nqm.irimager.IRImagerMock) → None

Stop video grabbing

class nqm.irimager.Logger

Bases: pybind11_object

Handles converting C++ logs to Python logging.Logger.

After you instantiate an object of this class, all spdlogs will no longer be printed to stderr. Instead, they’ll go to callback you’ve defined, or a logging.Logger.

Additionally, evo::IRLogger logs will also be captured.

Only a single instance of this object can exist at a time. You must destroy existing instances to create a new instance.

class nqm.irimager.LoggerContextManager

Bases: pybind11_object

Context Manager around a Logger object.

Designed for use with Python’s with syntax.

nqm.irimager.monotonic_to_system_clock(arg0: datetime.timedelta) → datetime.datetime

Converts from steady_clock to system_clock.

Converts a time_point from std::chrono::steady_clock (time since last boot) to std::chrono::system_clock (aka time since UNIX epoch).

C++20 has a function called std::chrono::clock_cast that will do this for us, but we’re stuck on C++17, so instead we have to do this imprecise monstrosity to do the conversion.

Remark:

s This function is imprecise!!! Calling it multiple times with the same data will result in different results.

Warning

The monotonic/steady_clock might only count when the computer is powered on. E.g. if the system was in a sleep state, the monotonic time may not have increased. Because of this, you should not rely on this function to return accurate results for past time points.