VideoCapture.read() using OpenCV. Why is USB Camera faster than CSI Camera? (NO GPU)

I am using NanoPi Duo 2 for the real-time image acquisition project.

I have noticed a significant performance (speed) difference between using CSI camera and USB camera.

The performance difference is as follows, Time for OpenCV VideoCapture.read()

CSI_OV5640_Camera = ~0.04s (40 ms) USB_Logitech_HD_C270 = ~0.009 (9 ms) 

As far as I know, I understand that NanoPi Duo2 does not have any GPU and CSI camera will be handled by CPU (Same as USB Camera).

Using $ htop both CSI and USB cameras shows 100% @ one of 4 cores.

For background,

OpenCV 3.4.6 Build Output

Video I/O  - libv4l/libv4l2    NO  - v4l/v4l2          linux/videodev2.h 

$ v4l2-ctl –get-fmt-video

Format Video Capture:         Width/Height      : 640/480         Pixel Format      : 'YV12'         Field             : Any         Bytes per Line    : 960         Size Image        : 460800         Colorspace        : Default         Transfer Function : Default         YCbCr Encoding    : Default         Quantization      : Default         Flags             : 

One interesting thing Performance of using numpy ndarray from CSI and USB are ALSO different

pyzbar.decode() calculation CSI_OV5640_Camera = ~0.43s (430 ms) USB_Logitech_HD_C270 = ~0.19s (190 ms) 

I am thinking that encoding of the VideoCapture.read() are different but all I see is same sized (640×480, 3 colors) numpy ndarray with similar looking values.

Thank you for reading my question.

Does USB read-rate from a camera device degrade when running inside of a docker container?

I have a general question regarding usb device performance inside of a docker container.

For context, I have setup a docker image (for deploying containers on an nvidia jetson agx xavier platform) to set up a camera publisher in ros to stream data from a realsense d435i camera plugged into the jetson via USB 3.2. I run this container with the --privileged option, as well as mounting the host volume /dev via -v /dev:/dev in the run command and I am able to bring up the node and run pretty reliably @15 fps at the camera’s lowest resolution. Additionally, I am able to view the camera stream on a separate host machine (connected over wlan) and the output looks pretty good.

I am trying to identify potential reasons why I cannot increase the frame-rate or the resolution without causing my ros node to crash. One of the issues, potentially, is the read-rate of my usb device when being read from within a running container. I am new to docker, and I would appreciate any general info and/or links about how running a docker container on a host machine can impact usb performance. Any relevant info would be of great use to me. Thanks!

Where is the epipole if one camera center is not in view of the other?

In the book multiple view geometry, the epipole is defined as follows:

The epipole is the point of intersection of the line joining the camera centres (the baseline) with the image plane. Equivalently, the epipole is the image in one view of the camera centre of the other view. It is also the vanishing point of the baseline (translation) direction.

The image of one camera center could easily not be in view of another camera center without the cameras being parallel. So, in this case, the line connecting the camera centers would not intersect at least one of the image planes. So, where is the epipole in this case?

Is a cryptographically signing camera possible?

What kind of digital / cryptographic signing would make it possible to create a camera that cryptographically signs the photos it takes so that the images can be verified to have been taken with that specific camera?

The motivation is that if the camera is able to implement a trust system, then it would create the ability to trust that photos or videos that it creates have not been subsequently modified.

Some potential variables:

  • Would it be preferred to use a single signing key for all cameras of a specific model, vs. keys for each instance of the camera, and how to limit the ability to extract/steal those keys. I guess this might involve a Trusted Platform Module, but I’m not knowledgeable about hardware security.

  • Out of scope for this question: The trust would be limited by the ability to present false image data to the camera at a stage prior to the signing step, using hardware hacking or simply holding up another image in front of the sensor.

  • Instead of implementing this as hardware, implement it as a smartphone app and take advantage of existing platform security.

After Factory Reset, Android Galaxy Camera Touchscreen is unresponsive

I met the weirdest thing I ever have. I have a galaxy camera gc100, It works very well without any problems. Since I decide to lend it to my friend, I did a factory reset. After the factory reset, everything looks fine and the initial setup screen “Select language” appears. BUT, I can not touch anything on the screen. It seems that the touch is simply unresponsive at all until the screen turns off by timeout. And then I tried (1) hard reboot, (2) go to Maintenance boot mode by pressing power and zoom out and did delete cache and factory reset again. (3) flash stock rom. (4) take out of the battery and then holding pressing the power button for more than 1 minute and then insert the battery and boot. (did three times). But none of them works. This is just weird and I can not find anyone on the Internet having the same issue as me. Other people met unresponsive touch screen and then did hard reset to fix the issue. Mine is entirely opposite. I don’t have any problems, but factory/hard reset causing the touch screen is unresponsive at initial setup screen. Do anyone have any clue how I can workaround this issue? Thanks a lot. (Btw the system is still functioning because I can adjust volume and take screenshots by using the physical buttons).

No camera app found in settings

I recently, factory resetted by android phone and installed my messenger. I tried videi calling my friend but i can’t switch from back to front cam, it says “couldn’t switch camera”. So i tried opening my default camera, it still works but i can’t find the switch camera button, even in other camera apps. I thought that i disabled my camera so i checked it in settings. I went to settings > apps > all apps but i can’t find it anywhere in the list. Even in the very button the list where disabled apps are located. I’m sure i have a front cam cause i video call/selfie before.

Desenvolvimento Android Câmera – Vídeo [pendente]

Preciso usar a câmera no Android para realizar gravação de vídeos, porém a câmera deve ser a pagina inicial do aplicativo, como no Snapchat, eu criei um protótipo como mostrado abaixo: Protótipo Tela Inicial

Estou começando no desenvolvimento Android e queria ajuda para saber se posso fazer isso via Java no Android Studio ou tem alguma maneira mais rápida e fácil, pois preciso para um projeto de TCC e meu tempo está curto.

Problem with making camera follow player in Javascript

So I’m testing a few things in Javascript and something I was testing was making the camera follow the player. When I move the player, instead of the dots going away from the player, they follow the player and it looks very strange.

Here is the code: https://codepen.io/anon/pen/wLgEPO?editors=0010

Move the player using the arrow keys.

Thanks.

Drone Thermal Camera price

Thermal Infrared Detector
Thermal infrared detector is a thermal imaging camera specifically tailored for small drones to meet all your thermal imaging requirements.
Product Description
Thermal infrared detector is a product with high performance.
▪It equipped with the drone does not require any illumination, which has a night vision function indeed, and can penetrate harsh conditions such as smoke and smog.
▪The video settings have specifically optimized for flight operations
▪Expansion needs in the future have been taken into consideration.
Specifications
OVERVIEW
Precision Mounting HolesTwo M2x0.4 on each of two sides & bottom One 1/4-20 threaded hole on top
Size2.26″ x 1.75″ (including lens)
Spectral Band7.5 – 13.5 µm
Thermal ImagerUncooled VOx Microbolometer
Weight3.25 – 4 oz (Configuration Dependant)
ZoomYes – Adjustable in App and via PWM
CONNECTIONS & COMMUNICATIONS
Analog Video OutputYes
POWER
Input Voltage4.8 – 6.0 VDC
Power Dissipation [peak]2.1 W (3.9 W)
Feature
▶Small size and light weight. Such structure design can minimize the influence on aircraft weight, drone stability and flight time.
▶High efficiency. The detector is able to catch the whole image and also get all-over thermal information in real-time.
Application
In the field of public security
In order to arrest the suspects, it cost a lot of financial and manpower resources since dim light and complex environment often increase difficulty in finding targeted suspects.
▶Now, with the help of drones equipped with thermal infrared detector, it has greatly reduced the cost of police search and reduced the time of search and has become an important modern equipment for police to crack down on crime at night.Drone Thermal Camera price
website:http://www.rdroneguy.com/drone-thermal-camera/