IQ.Pilot Prebuilt Release @ 27f668a
This commit is contained in:
217
iqpilot/system/camerad/test/test_camerad.py
Normal file
217
iqpilot/system/camerad/test/test_camerad.py
Normal file
@@ -0,0 +1,217 @@
|
||||
import os
|
||||
import time
|
||||
|
||||
import pytest
|
||||
import numpy as np
|
||||
from msgq.visionipc import VisionIpcClient
|
||||
|
||||
import iqpilot.cereal.messaging as messaging
|
||||
from iqpilot.cereal.services import SERVICE_LIST
|
||||
from iqpilot.selfdrive.test.helpers import processes_context
|
||||
from iqpilot.system.camerad.snapshot import VISION_STREAMS
|
||||
from iqpilot.system.manager.process_config import managed_processes
|
||||
from iqpilot.tools.lib.logreader import msgs_to_time_series
|
||||
|
||||
TEST_TIMESPAN = 10
|
||||
CAMERAS = ('roadCameraState', 'driverCameraState', 'wideRoadCameraState')
|
||||
TEST_PATTERN_FRAMES = 200
|
||||
TEST_PATTERN_MIN_CONFIDENCE = 10
|
||||
TEST_PATTERN_CONNECT_TIMEOUT = 15
|
||||
TEST_PATTERN_CONFIGS = {
|
||||
'ox03c10': (41, 4),
|
||||
'os04c10': (97, 4),
|
||||
}
|
||||
|
||||
|
||||
def _pattern_sample(client):
|
||||
buf = client.recv(1000)
|
||||
if buf is None:
|
||||
return None
|
||||
|
||||
y = np.asarray(buf.data[:buf.uv_offset], dtype=np.uint8).reshape((-1, buf.stride))[:buf.height, :buf.width]
|
||||
profile = y[:, ::8].mean(axis=1)
|
||||
padded = np.pad(profile, (4, 4), mode='edge')
|
||||
neighbors = [padded[i:i + len(profile)] for i in range(9) if i != 4]
|
||||
residual = profile - np.median(neighbors, axis=0)
|
||||
position = int(np.argmax(residual))
|
||||
return client.frame_id, client.timestamp_sof, position, residual[position], buf.height
|
||||
|
||||
|
||||
def _test_pattern_session():
|
||||
samples = {camera: [] for camera in CAMERAS}
|
||||
sockets = {camera: messaging.sub_sock(camera, conflate=False, timeout=100) for camera in CAMERAS}
|
||||
logs = []
|
||||
with pytest.MonkeyPatch.context() as monkeypatch:
|
||||
monkeypatch.setenv('SPECTRA_TEST_PATTERN', '1')
|
||||
monkeypatch.setenv('SPECTRA_ERROR_PROB', '-1')
|
||||
with processes_context(['camerad']) as processes:
|
||||
clients = {camera: VisionIpcClient('camerad', VISION_STREAMS[camera], False) for camera in CAMERAS}
|
||||
pending = set(clients)
|
||||
deadline = time.monotonic() + TEST_PATTERN_CONNECT_TIMEOUT
|
||||
while pending and time.monotonic() < deadline:
|
||||
assert processes[0].proc is not None and processes[0].proc.exitcode is None
|
||||
pending = {camera for camera in pending if not clients[camera].connect(False)}
|
||||
if pending:
|
||||
time.sleep(0.1)
|
||||
assert not pending, f'VisionIPC connection timeout: {sorted(pending)}'
|
||||
|
||||
for _ in range(TEST_PATTERN_FRAMES):
|
||||
for camera, client in clients.items():
|
||||
sample = _pattern_sample(client)
|
||||
if sample is not None:
|
||||
samples[camera].append(sample)
|
||||
for sock in sockets.values():
|
||||
logs.extend(messaging.drain_sock(sock))
|
||||
|
||||
return msgs_to_time_series(logs), samples
|
||||
|
||||
|
||||
def run_and_log(procs, services, duration):
|
||||
logs = []
|
||||
|
||||
try:
|
||||
for p in procs:
|
||||
managed_processes[p].start()
|
||||
socks = [messaging.sub_sock(s, conflate=False, timeout=100) for s in services]
|
||||
|
||||
start_time = time.monotonic()
|
||||
while time.monotonic() - start_time < duration:
|
||||
for s in socks:
|
||||
logs.extend(messaging.drain_sock(s))
|
||||
for p in procs:
|
||||
assert managed_processes[p].proc.is_alive()
|
||||
finally:
|
||||
for p in procs:
|
||||
managed_processes[p].stop()
|
||||
|
||||
return logs
|
||||
|
||||
@pytest.fixture(scope="module")
|
||||
def logs():
|
||||
logs = run_and_log(["camerad", ], CAMERAS, TEST_TIMESPAN)
|
||||
ts = msgs_to_time_series(logs)
|
||||
|
||||
for cam in CAMERAS:
|
||||
expected_frames = SERVICE_LIST[cam].frequency * TEST_TIMESPAN
|
||||
cnt = len(ts[cam]['t'])
|
||||
assert expected_frames*0.8 < cnt < expected_frames*1.2, f"unexpected frame count {cam}: {expected_frames=}, got {cnt}"
|
||||
|
||||
dts = np.abs(np.diff([ts[cam]['timestampSof']/1e6]) - 1000/SERVICE_LIST[cam].frequency)
|
||||
assert (dts < 1.0).all(), f"{cam} dts(ms) out of spec: max diff {dts.max()}, 99 percentile {np.percentile(dts, 99)}"
|
||||
return ts
|
||||
|
||||
@pytest.mark.tici
|
||||
class TestCamerad:
|
||||
def test_frame_skips(self, logs):
|
||||
for c in CAMERAS:
|
||||
assert set(np.diff(logs[c]['frameId'])) == {1, }, f"{c} has frame skips"
|
||||
|
||||
def test_frame_sync(self, logs):
|
||||
n = range(len(logs['roadCameraState']['t'][:-10]))
|
||||
|
||||
frame_ids = {i: [logs[cam]['frameId'][i] for cam in CAMERAS] for i in n}
|
||||
assert all(len(set(v)) == 1 for v in frame_ids.values()), "frame IDs not aligned"
|
||||
|
||||
frame_times = {i: [logs[cam]['timestampSof'][i] for cam in CAMERAS] for i in n}
|
||||
diffs = {i: (max(ts) - min(ts))/1e6 for i, ts in frame_times.items()}
|
||||
|
||||
laggy_frames = {k: v for k, v in diffs.items() if v > 1.1}
|
||||
assert len(laggy_frames) == 0, f"Frames not synced properly: {laggy_frames=}"
|
||||
|
||||
def test_sanity_checks(self, logs):
|
||||
self._sanity_checks(logs)
|
||||
|
||||
def _sanity_checks(self, ts):
|
||||
for c in CAMERAS:
|
||||
assert c in ts
|
||||
assert len(ts[c]['t']) > 20
|
||||
|
||||
# not a valid request id
|
||||
assert 0 not in ts[c]['requestId']
|
||||
|
||||
# should monotonically increase
|
||||
assert np.all(np.diff(ts[c]['frameId']) >= 1)
|
||||
assert np.all(np.diff(ts[c]['requestId']) >= 1)
|
||||
|
||||
# EOF > SOF
|
||||
assert np.all((ts[c]['timestampEof'] - ts[c]['timestampSof']) > 0)
|
||||
|
||||
# logMonoTime > SOF
|
||||
assert np.all((ts[c]['t'] - ts[c]['timestampSof']/1e9) > 1e-7)
|
||||
|
||||
# logMonoTime > EOF, needs some tolerance since EOF is (SOF + readout time) but there is noise in the SOF timestamping (done via IRQ)
|
||||
assert np.mean((ts[c]['t'] - ts[c]['timestampEof']/1e9) > 1e-7) > 0.7 # should be mostly logMonoTime > EOF
|
||||
assert np.all((ts[c]['t'] - ts[c]['timestampEof']/1e9) > -0.10) # when EOF > logMonoTime, it should never be more than two frames
|
||||
|
||||
def test_stress_test(self):
|
||||
os.environ['SPECTRA_ERROR_PROB'] = '0.008'
|
||||
logs = run_and_log(["camerad", ], CAMERAS, 10)
|
||||
ts = msgs_to_time_series(logs)
|
||||
|
||||
# we should see some jumps from introduced errors
|
||||
assert np.max([ np.max(np.diff(ts[c]['frameId'])) for c in CAMERAS ]) > 1
|
||||
assert np.max([ np.max(np.diff(ts[c]['requestId'])) for c in CAMERAS ]) > 1
|
||||
|
||||
self._sanity_checks(ts)
|
||||
|
||||
|
||||
@pytest.fixture(scope="module")
|
||||
def test_pattern_data():
|
||||
return _test_pattern_session()
|
||||
|
||||
|
||||
@pytest.mark.tici
|
||||
@pytest.mark.xdist_group("camerad_test_pattern")
|
||||
class TestCameradTestPattern:
|
||||
def test_frame_delivery(self, test_pattern_data):
|
||||
logs, samples_by_camera = test_pattern_data
|
||||
for camera in CAMERAS:
|
||||
assert camera in logs
|
||||
samples = samples_by_camera[camera]
|
||||
assert len(samples) > TEST_PATTERN_FRAMES * 0.9
|
||||
|
||||
state_frame_ids = logs[camera]['frameId']
|
||||
state_request_ids = logs[camera]['requestId']
|
||||
vipc_frame_ids = np.array([sample[0] for sample in samples])
|
||||
for source, frame_ids in (('camera state', state_frame_ids), ('VisionIPC', vipc_frame_ids)):
|
||||
frame_steps = np.diff(frame_ids)
|
||||
skipped = frame_ids[1:][frame_steps != 1]
|
||||
assert len(skipped) == 0, f'{camera} {source} skipped frames before {skipped}'
|
||||
|
||||
expected_sof_step = 1e9 / SERVICE_LIST[camera].frequency
|
||||
sof_step_errors = np.diff(logs[camera]['timestampSof']) - expected_sof_step
|
||||
assert np.all(np.abs(sof_step_errors) < 1e6), f'{camera} SOF cadence errors: {sof_step_errors[np.abs(sof_step_errors) >= 1e6]}'
|
||||
|
||||
request_steps = np.diff(state_request_ids)
|
||||
skipped_requests = state_request_ids[1:][request_steps != 1]
|
||||
assert len(skipped_requests) == 0, f'{camera} skipped requests before {skipped_requests}'
|
||||
|
||||
state_sofs = dict(zip(state_frame_ids, logs[camera]['timestampSof'], strict=True))
|
||||
matched_samples = [sample for sample in samples if sample[0] in state_sofs]
|
||||
assert len(matched_samples) > len(samples) * 0.8
|
||||
mismatched_sofs = {
|
||||
frame_id: (timestamp_sof, state_sofs[frame_id]) for frame_id, timestamp_sof, *_ in matched_samples if timestamp_sof != state_sofs[frame_id]
|
||||
}
|
||||
assert not mismatched_sofs, f'{camera} VisionIPC/camera state SOFs disagree: {mismatched_sofs}'
|
||||
|
||||
def test_pattern(self, test_pattern_data):
|
||||
logs, samples_by_camera = test_pattern_data
|
||||
for camera in CAMERAS:
|
||||
sensors = set(logs[camera]['sensor'])
|
||||
assert len(sensors) == 1
|
||||
sensor = sensors.pop()
|
||||
assert sensor in TEST_PATTERN_CONFIGS, f'unsupported test pattern sensor: {sensor}'
|
||||
cycle_frames, position_tolerance = TEST_PATTERN_CONFIGS[sensor]
|
||||
|
||||
samples = samples_by_camera[camera]
|
||||
confident = [sample for sample in samples if sample[3] > TEST_PATTERN_MIN_CONFIDENCE]
|
||||
positions = np.array([sample[2] for sample in confident])
|
||||
assert len(confident) > len(samples) * 0.7, f'{camera} test pattern confidence too low'
|
||||
assert len(np.unique(positions)) > 20, f'{camera} test pattern is not moving'
|
||||
assert np.ptp(positions) > confident[0][4] * 0.75, f'{camera} test pattern does not span the frame'
|
||||
|
||||
samples_by_frame = {sample[0]: sample for sample in confident}
|
||||
repeating_pairs = [(sample, samples_by_frame[sample[0] + cycle_frames]) for sample in confident if sample[0] + cycle_frames in samples_by_frame]
|
||||
assert len(repeating_pairs) > 20
|
||||
unexpected = [(first[0], first[2], second[2]) for first, second in repeating_pairs if abs(second[2] - first[2]) > position_tolerance]
|
||||
assert len(unexpected) < len(repeating_pairs) * 0.3, f'{camera} test pattern cycle mismatches: {unexpected}'
|
||||
Reference in New Issue
Block a user