841 Reply by joan

(3 replies, posted in Cameras and hardware)

Very good point about multiple view analysis. It departs a bit from what I thought with video quality-degrading issues, but I like to consider what the imaging system can deliver as a whole, be it a single or multiple camera system.

For 3D quantitative analysis, I think 2 cameras is the theoretical minimum but that real motion tracking applications never use less than 4 cameras. (I've used 2-camera systems for tracking eyes or fingers in a small volume, but as soon as you want to track body parts you move to 4 to 8 or even more).

So, what can constitute the main problems and defects of a multi-camera setup ? (Considering qualitative analysis only for now).
Mis-synchronization is probably one of the biggest… We can always synchronize to frame-level in software, but sub-frame sync requires hardware support. That's a clever use of WiFi for sure.
I don't know what constitute an acceptable synchronization level for sport analysis ? (I know that stereoscopy video requires almost pixel level sync for example).
On the subject, I have plans for a half software, half hardware sub-frame level synchronization method using the rolling shutter on consumer USB cameras and an Arduino powered strobe light, I'll post more details if I ever get it working.

Go to forum: Cameras and hardware

842 Reply by joan

(3 replies, posted in Cameras and hardware)

Resolving power - lack of focus
- Impacts: the sharpness of details.
- Relevance to sport analysis: high.
- Component: Lens and lens mount.
- How to control or improve:
If the subject is always distant, a fixed focus camera may be sufficient. A camera with fixed focus should hopefully be focused at infinity in factory.
The most versatile solution is a manual focus that can be adjusted with a ring or lever.
The most efficient solution may be a motorized focus that we can control in software. (Logitech C920, Microsoft LifeCam). Note that even with motorized focus some webcams can't focus to infinity and anything farther than a few meters, will not be optimally focused.
Some lenses have variable focal length, in this case focusing sould usually be redone after changing the focal length.
Some devices have auto-focus capabilities, in this case care should be taken as to where in the image the focus has been locked.

Resolving power - long exposure
- Impacts: the sharpness of details on moving subjects.
- Relevance to sport analysis: very high.
- Component: Sensor.
- How to control or improve:
Some cameras have auto-exposure, they will adjust exposure to measured light levels. It lower reproducibility and the final exposure choosen may not be adequate (long exposure increases motion blur).
The most versatile solution is a camera for which exposure duration can be changed manually and is capable of short exposures (Exact requirement to be assessed).
- Compromise: low exposure means less light collected at the pixel sites. For laboratory setups artificial lights may be needed.

Resolving power - pixel count and lens resolution
- Impacts: the sharpness of details.
- Relevance to sport analysis: high.
- Component: Sensor and lens.
- How to control or improve:
Some devices are actually limited by their lens, when the lens itself is not able to project an image sharp enough to distinguish details that are two pixels apart.
More pixels is better but only if the lens is adequate. For a given sensor size, more pixels means smaller ones, which makes it more difficult for the lens to match resolution.
Lenses quality is measured in various metrics like lp/ph or MTF curves. A recent evolution is the use of Megapixel ratings. The lens fitted on the camera should have a megapixel rating at least as high as the pixel count of the sensor. (ex: A 3MP rated lens for 1920x1080 images). A good introductory resource on lens quality measurement methodology is at Cambridge in Colour.

Resolving power - image processing and JPEG compression
- Impacts: the sharpness of details.
- Relevance to sport analysis: high.
- Component: Image processing chip on the camera or recording software.
- How to control or improve:
The best solution is a camera that can provide the raw images to the computer, and to perform the color grading there.
The issue is that bandwidth is limited, so it is not always possible to transmit full color frames at the full framerate.
A camera should allow us to control the JPEG compression levels. (No USB camera currently does this to my knowledge).

Spherical distortions - wide angle and ultra wide angle lenses
- Impacts: measurements of distances and speeds.
- Relevance to sport analysis: mid to high.
- Component: Lens.
- How to control or improve:
Lenses with normal field of view (less than around 65°) usually have very low distortion.
For wide angle, a lens without distortions should be preferred, but the cost can skyrocket pretty quickly.
The distortion can be calibrated in software and taken into account for measurements.
- Compromise: A subject evolving at the same distance from the camera will cover less pixels, so less resolution.

Go to forum: Cameras and hardware

843 Topic by joan

(3 replies, posted in Cameras and hardware)

I would like to use this thread to compile a list of quality-degrading factors in video, how much they are relevant to sport analysis, which component are involved, and how we may improve upon them.

This list should be general and relevant to anything that can provide a stream of image and store it on-device or transmit it to a computer (DSLR, Industrial camera, USB camera, IP camera, Smartphone, etc.). The trigger for this topic though, is the advent of high-quality, interchangeable, small lenses for surveillance-type cameras. We are now very near a day where little USB cameras can be considered serious imaging devices.

Please add your input, illustrative images, comments, remarks, additional degrading factors, formatting suggestions, etc.
Maybe at some point we can create a PDF or something. It should be useful for evaluating new hardware on the market and as a buyers guide.

Here are some topics that could be covered:

  • limited resolving power - lack of focus.

  • limited resolving power - long exposure.

  • limited resolving power - pixel count and lens resolution.

  • limited resolving power - image processing and JPEG compression.

  • spherical distortions - wide and ultra wide angle lenses.

  • vignetting - mechanical and optical.

  • noise.

  • flares.

  • limited temporal sampling granularity - low framerate.

  • temporal distortion - rolling shutter.

  • limited illuminance - low aperture.

  • limited dynamic range.

  • limited depth of field.

  • chromatic aberrations.

  • unfaithful color reproduction.

Go to forum: Cameras and hardware

844 Reply by joan

(4 replies, posted in General)

Ok, this is officially very cool :-)
I wonder why I never tried that, thank you for the heads up!

Exposure compensation seems to change the gain level. I could not find a setting for exposure duration unfortunately. The auto-exposure is what degrades the framerate in low-light.

It made me realize that the image size should be displayed in Kinovea for these streams, as it's not always given.

Go to forum: General

845 Reply by joan

(4 replies, posted in General)

Hi,
I have not yet tried to serve smartphone camera stream as MJPEG. It is quite relevant to the current effort in the capture module.
What application do you use on the device and what type of framerate/frame size does it achieve ?

Go to forum: General

846 Reply by joan

(2 replies, posted in General)

Hi,
The second issue is linked to the first. Kinovea will mark the file as 30fps. If the incoming stream was actually 9fps it will be played back accelerated.

There are several things that can cause a decrease in stream framerate. The easiest one is when the exposure duration is set to longer than the frame interval. Exposure will take precedence and the framerate will be lowered automatically and silently.

Go into the settings, device properties, locate the exposure parameter, make sure "Auto" in not checked, and slide to the left. The image will get darker and darker as you decrease the duration of exposure, this is expected. Test with various settings to see the impact on framerate.

The next version of Kinovea will allow you to get more out of the cameras. The Microsoft LifeCam Studio has MJPEG on-board compression and Kinovea can leverage it to improve recording performances.

Go to forum: General

847 Reply by joan

(7 replies, posted in General)

Hi,
I have not, but my development machine is now x64. I hope to be able to work on this and have something out sometime this year.

Go to forum: General

848 Reply by joan

(12 replies, posted in Cameras and hardware)

You should have something roughly like this :

http://www.kinovea.org/screencaps/0.8.x/0824-graphstudionext.png

The right click must be done on the tiny square to the right of "Capture", but I guess you already found that if you rendered the pin.

Go to forum: Cameras and hardware

849 Reply by joan

(12 replies, posted in Cameras and hardware)

This camera module is quite interesting : ELP-USBFHD01M from Ailipu Technology. (You may find it on your favorite Chinese reseller at about 35€ + shipping).

It is based on an Omnivision OV2710 sensor which does 1920x1080@30fps, 1280x720@60fps and 640x480@120fps.
They apparently implemented MJPEG compression on all of these sensor outputs and the camera is UVC compliant.
It comes with various M12 lenses options so the lens might be interchangeable.
It is not clear whether it has manual exposure or not. I just ordered some to find out if it delivers on the specs.

Go to forum: Cameras and hardware

850 Reply by joan

(27 replies, posted in Cameras and hardware)

Re-lenses, I think the C920 have a proprietary lens mount and the lens can't be swapped easily. (Haven't disassembled mine yet).

One thing that crossed my radar is this S-Mount for the C920 board.

Ah, if only all USB cameras used standard S-mount so we could swap in M12 lenses that would be great…
There are high quality wide-angle M12 lenses that would be very interesting to use, and we could reuse them when upgrading camera.

This guy went the other way and created a CS-mount adapter.

Go to forum: Cameras and hardware

851 Reply by joan

(12 replies, posted in Cameras and hardware)

I have also yet to find a 60fps USB 2.0 camera with more than 640x480 resolution.
- The PS3Eye does 640x480 @ 75fps.
- The logitech C910 is apparently also capable of 640x480 @ 60 fps. The option was removed in the C920 for some reason.
- The C920-c (business version?) is reported by some sources as 960x720 @ 60 fps but I can't verify.
- I don't know about the C930 or C615.
- There is also a Chinese clone Gucee that claims 640x480 @ 60 fps.

1080p @ 60 fps is nowhere to be found…

It shouldn't be a bandwidth problem for the recent Logitech cameras like the C920 because unlike the PS3Eye and other cameras they have on-board MJPEG and H.264 compression. Maybe the compression chip is not fast enough to keep up, or they just didn't bother with it as it's less likely to interest their primary market.

The next options are USB 3.0 and GigE cameras, but the price jump is painful. Also the stream is raw so now it's the PC-side and the hard-drive that have to cope with compression and writing speed.

Go to forum: Cameras and hardware

852 Reply by joan

(12 replies, posted in Cameras and hardware)

One approach to streamline the workflow without live streaming would be to use a WiFi-enabled SD card. Ideally we would like to access the card as a network drive…

I haven't tried these in a while so I can't recommend a specific product. The original Eye-Fi had some workflow limitations that made it a bit frustrating to use. I don't know if they improved upon this.

I do not know if the camera can stay on while the WiFi-enabled part of the card is reading the files. There might be sharing issues at the filesystem level preventing this.

If someone has experimented with these please share.

The GoPro also has some WiFi options and an embedded HTTP server. I don't know if anyone ever found out how to access SD cards files through this way.

Go to forum: Cameras and hardware

853 Reply by joan

(12 replies, posted in Cameras and hardware)

slim_n_trim wrote:

So that we can open files on the GoPro via Kinovea directly (live capturing).

Opening files is different from live capture. Opening files should be no problem, the SD card is seen as a storage device.

I have no direct experience with the cable you mention and haven't passed the page through translation, however note that a cable for FPV is bound to be analog (to limit latency) and will connect to a radio transmitter rather than a PC.

Other video out options that I know of on the GoPro are the small preview stream over WiFi (not supported in Kinovea) and the HDMI out which needs an additional dedicated capture hardware (of which I do not know if they are well supported in Kinovea or not).

Go to forum: Cameras and hardware

854 Reply by joan

(1 replies, posted in Bug reports)

Yes you should be able to do that.
Move to the wanted position with the arrow keys and hit the working zone buttons (green square brackets).

Go to forum: Bug reports

855 Reply by joan

(2 replies, posted in Bug reports)

First of all I'm very sorry for the lost file. Let's get to the bottom of this so that doesn't happen again.

I'll try to summarize my understanding of the issues, please correct me if I'm wrong.

1. Your workflow is to record video pairs with two Sony DSC-H20 using the internal memory cards. Then you transfer the files manually to the computer. (No live capture).
2. You record a single master video for each camera.
3. For each camera master video, you then create several sub sequences. You do this using Kinovea, setting a working zone and saving it.

Issue 1: When reopening a pair of sub-sequences, the endpoint and duration do not match, although the original working zones matched.

4. For each pair of sub-sequence, you open the files in Kinovea, create a new working zone to fix the endpoint from issue 1.
5. You save the fixed-up sub-sequences again.

Issue 2: When saving one of the video, the other video file becomes corrupted.
Issue 3: QuickTime cannot open any video after it has passed through Kinovea saving.

Please correct any misunderstanding.

Let's also name things:
1. Creation of master files A and B in the cameras.
2. Creation of files A1, A2, …, A12, and B1, B2, …, B12 in Kinovea.
3. Creation of files A1fix, A2fix, etc. and B1fix, B2fix, etc. in Kinovea.

The major issue is #2, file corruption.
There is one critical bug that can cause this, it happens when you overwrite an open file. For example, if you open A12 and B12, fix A12 endpoint, then in the Save File dialog you either save it back to A12 or B12. Please create a backup of your files before attempting a reproduction of the issue.

I'll add the necessary check to prevent this failure.

edit: analysing the log, I think it is indeed what happened:

1673000 - DEBUG - [Main] - VideoReaderFFMpeg - [File] - Filename : Part11-030-Oil.MP4
...
1683468 - DEBUG - [Main] - VideoReaderFFMpeg - [File] - Filename : Part11-462-H2O.MP4
...
1802000 - DEBUG - [Saving] - FrameServerPlayer - Saving selection [1001]->[225225] to: Part11-030-Oil.MP4
...
1823468 - DEBUG - [Saving] - FrameServerPlayer - Saving selection [1001]->[238238] to: Part11-462-H2O.MP4

Go to forum: Bug reports