Resizing camera images

[Okeskin]

Hi everyone,

I have a screen project displaying a camera image using an STM32N6. I've validated and run several peripherals like DCMIPP and DMA2D. I'm currently displaying a 720x576 resolution image from DCMIPP on the screen via layer 2 using bitmap addressing.

My question is: Since I'm using internal RAM, Touchgfx uses one layer, and one layer is used for the camera. Because I'm not processing any bitmaps and need space, I've significantly reduced the size of the cmd_list area. After adding all this, I have about 300kB of free RAM left.

Now, what I need is to be able to see this camera image in full size on my 1024x600 px screen. However, I want to know if this enlargement is possible and if it requires a cmd_list, i.e., vector graphics.

Will I use vector graphics for this? Will GPU2D use the RAM allocated for layer 2 for this operation, or will it require an additional amount of RAM equal to the screen size?

 

Best Regards,

Onur

 

[mathiasmarkussen]

Hello Okeskin,

 

It should be possible for you to bind the camera buffer in RAM to a dynamic bitmap in TouchGFX, If you use that dynamic bitmap as the input to a resizable image or a texture mapper, you can show the image in TouchGFX and scale it from there. We have shared an example of how this can be done on the forums not too long ago.

I would generally not advise changing the cmdlist size. Have you set the layer framebuffer size of the camera LTDC layer to 720x576 pixels? that should be all you need.