CroatianMeteorNetwork
RMS stops image capturing during strong wind and camera vibration
Hi, I have camera inside large box and during strong wind some vibration is happening (as result blurry fits images ). But noticed if wind is very strong: RMS just stops image capturing without any error in log and you just missed the most of nigh.
It was reproduced 3rd time over 5 months . Example of log below:
2024/01/11 22:36:55-INFO-BufferedCapture-line:474 - New block of raw frames available for compression with starting time: 1705012595.1614099
2024/01/11 22:36:55-INFO-BufferedCapture-line:331 - Grabbing a new block of 256 frames...
2024/01/11 22:36:55-DEBUG-Compression-line:274 - Compressing frame block with start time at: 1705012595.1614099
2024/01/11 22:37:02-DEBUG-Compression-line:295 - Compression time: 7.219 s
2024/01/11 22:37:02-DEBUG-Compression-line:302 - Saving time: 0.068 s
2024/01/11 22:37:03-INFO-DetectionTools-line:61 - Loaded mask: mask.bmp
2024/01/11 22:37:03-DEBUG-Compression-line:319 - Extractor started for: UA0004_20240111_223635_161_0110336
2024/01/11 22:37:03-INFO-DetectStarsAndMeteors-line:63 - Running detection on file: FF_UA0004_20240111_223635_161_0110336.fits
2024/01/11 22:37:03-INFO-DetectionTools-line:61 - Loaded mask: mask.bmp
2024/01/11 22:37:03-INFO-Compression-line:332 - Added file for detection: FF_UA0004_20240111_223635_161_0110336.fits
2024/01/11 22:37:03-DEBUG-VideoExtraction-line:285 - [UA0004_20240111_223635_161_0110336] time for thresholding and subsampling: 0.4954211711883545s
2024/01/11 22:37:03-DEBUG-VideoExtraction-line:289 - [UA0004_20240111_223635_161_0110336] nothing found, not extracting anything 1
2024/01/11 22:37:05-INFO-ExtractStars-line:186 - extracted 39 stars from FF_UA0004_20240111_223635_161_0110336.fits
2024/01/11 22:37:05-INFO-DetectStarsAndMeteors-line:89 - Detected stars: 1
2024/01/11 22:49:24-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/11 23:19:25-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/11 23:49:25-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 00:19:26-INFO-EventMonitor-line:1067 - Added event at 20231228_054329 to the database
2024/01/12 00:19:26-INFO-EventMonitor-line:1775 - Checks on trajectories for event at 20231228_054329
2024/01/12 00:19:26-INFO-EventMonitor-line:1483 - Using .config as .config file name
2024/01/12 00:19:26-INFO-EventMonitor-line:1785 - No files for event - marking 20231228_054329 as processed
2024/01/12 00:19:26-INFO-EventMonitor-line:1093 - Event at 20231228_054329 marked as processed
2024/01/12 00:19:26-INFO-EventMonitor-line:1921 - 1 event was processed, EventMonitor work completed
2024/01/12 00:19:26-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 00:49:27-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 01:19:27-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 01:49:28-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 02:19:28-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 02:49:29-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 03:19:30-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 03:49:31-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 04:19:31-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 04:49:32-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 05:19:33-INFO-EventMonitor-line:2007 - Event monitor check completed
2024/01/12 05:20:41-DEBUG-StartCapture-line:390 - Stopping capture...
2024/01/12 05:20:42-INFO-BufferedCapture-line:96 - Joining capture...
2024/01/12 05:21:42-INFO-BufferedCapture-line:106 - Terminating capture...
2024/01/12 05:21:42-DEBUG-StartCapture-line:393 - Capture stopped
The issue is probably a dropped connection to the camera or some other video issue. Have you tried connecting to the camera in the morning after failure?
Hi, camera connected directly to RPi (MIPI camera). According to the log record:
2024/01/12 05:21:42-INFO-BufferedCapture-line:106 - Terminating capture...
It stuck in main loop during grabbing frames and I found only one possible place for deadlock:
ret, frame = device.read()
So it could be opencv or hardware issue
In opencv version 4.8 added timeout properties : CAP_PROP_READ_TIMEOUT_MSEC
timeout in milliseconds for reading from a video capture (applicable for FFmpeg and GStreamer back-ends only)
Probably it might help for reconnection to camera
Tried to reproduce on new device - can't reproduce. With vibration, movements and hits still continue processing. Need check on original device, seems like hardware issue.