• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 103 subscribers
  • Views 139 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

TDA4VM: Using y_data and cbcr_data for pixel information

TDA4VM: resizeImage takes too long to run on target (c66)

Prithvi Kambhampati
Prodigy 235 points
Part Number: TDA4VM

Hi,

We are using the resizeImage function from app_tidl to upscale images in the target (c66) for app_tidl_od but we find that it adds more than 100ms latency to the entire pipeline.

1. Are there ways to optimize it?

2. Does this function upscale a smaller image (30x40pixels) to a larger image (256x256)? or does it just downscale?

Regards,

Prithvi

  • 0
    TI__Expert 4305 points

    HI,

    This is plain C code, and I guess it can do up scaling and downscaling both. Definitely it can be optimized using C intrinsics of C66x DSP. Theoretical compute performance for this loop should be ~3 cycle per output leaving the border corner scenario. Since in the demo resize resolution is low (256x256) hence cycle consumed is low and so plain C code is used there without any optimization. There are many online material available for C66x optimization tricks, you may refer that for optimizing this.

    Regards

    Deepak Poddar