haipnh

Purposes

1. iGPU will be used for main display and rendering daily-use softwares in default
2. Dedicated GPUs will be used for computing or offload rendering
3. Reduce the used VRAM of Dedicated GPU, which is mostly used for X11 server rendering

---

Tested Hardware Configuration

OS: Ubuntu 18.04.4 LTS
Mainboads
├── iGPU

haipnh

sudo -s
echo "source /etc/profile.d/pynq_venv.sh" >> ~/.bashrc
echo "export XILINX_XRT=/usr" >> ~/.bashrc
source ~/.bashrc
python some_script.py

haipnh

import numpy as np
import cv2
from IPython.display import display, Image
import time

video_path = "./HowDeepIsYourLove.mp4"

video = cv2.VideoCapture(video_path)
display_handle=display(None, display_id=True)
kernel = np.ones((5,5),np.float32)/25

haipnh

# Summary: We will create a virtual environment of Python-3.9 and install OpenCV-Python to it.

# 1. Add 3rd-party OpenCV-3.9 repository
sudo apt-key adv --refresh-keys --keyserver keyserver.ubuntu.com
sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt-get update

# 2. Install packages
sudo apt-get install -y software-properties-common python3-pip python3.9 python3.9-dev python3.9-venv

haipnh

#!/bin/bash

# This gist provides several test cases to verify a host machine using NVIDIA graphics card if it's ready for tensorflow and pytorch.

### TC#1: Verify hardware information
# Expectation: The OS shows hardware bare information
lspci | grep -i nvidia
sudo lshw -C display

### TC#2: Verify nvidia-driver-X

haipnh

libdir = /opt/opencv-4.0.0/lib
includedir = /opt/opencv-4.0.0/include/opencv4

Name: OpenCV
Description: OpenCV (Open Source Computer Vision Library) is an open source computer vision and machine learning software library.
Version: 4.0.0
Libs: -L${libdir} -lopencv_aruco -lopencv_bgsegm -lopencv_bioinspired -lopencv_calib3d -lopencv_ccalib -lopencv_core -lopencv_datasets -lopencv_dnn_objdetect -lopencv_dnn -lopencv_dpm -lopencv_face -lopencv_features2d -lopencv_flann -lopencv_freetype -lopencv_fuzzy -lopencv_gapi -lopencv_hfs -lopencv_highgui -lopencv_imgcodecs -lopencv_img_hash -lopencv_imgproc -lopencv_line_descriptor -lopencv_ml -lopencv_objdetect -lopencv_optflow -lopencv_phase_unwrapping -lopencv_photo -lopencv_plot -lopencv_reg -lopencv_rgbd -lopencv_saliency -lopencv_shape -lopencv_stereo -lopencv_stitching -lopencv_structured_light -lopencv_superres -lopencv_surface_matching -lopencv_text -lopencv_tracking -lopencv_videoio -lopencv_video -lopencv_videostab -lopencv_xfeatures2d -lopencv_ximgproc -lopenc

haipnh

This is a note on how to cross compile opencv for pretty much any ARM device(HardFP supported in this case) and deploy. Native
compiling in ARM devices can be painfully slow and they seem to hang often during build(mine got stuck at 43%). So if you happen
to have a desktop/laptop/server running ubuntu or similar linux distro, u can build opencv in fractionth of the time taken for 
native compiling without any issues.
Building opencv3 with TBB and NEON and VFP support can boost opencv performance. Thanks to Adrian at pyimagesearch for pointing that out.
Both my PC and target machine aka orange pi zero are running ubuntu 16.04 with python2.7 and python 3.5.

1.Run the following commands in both machines to install the necessary libraries etc.(mine worked with them,so they should be enough 
hopefully)
sudo apt-get update && sudo apt-get upgrade