I currently have a program which uses the camera, video_render and resize component, along with OpenGLES and egl_render. The goal is to display the fullHD feed on the main HDMI screen (the video_render component), and display a second smaller feed on the 7" touchscreen (with the egl_render component) on which we can then run face detection with OpenCV.
Basically, the component chain is like this :
camera (preview port 70) -> video_render (full HD)
camera (capture port 71) -> resize -> egl_render
Everything is working fine, but what I now need is a way to save an image in fullHD on demand. Basically a function which when called, would save the currently displayed image on the video_render (or the next image, it doesn't really matter) to a file.
I don't really know the best way to do this. Should I try to use the port 72 of the camera (apparently used to capture still images). Can I just intercept the buffer going into the video_render and save it somehow ? Is there an even simpler way ?