Tag Archives: Microsoft.Extensions.Configuration

IoT and AI Building Edge AI with AI- Introduction

Posted on by

Introduction

Recently, there have been mainstream media articles (Forbes, Wired-Video games , Wired-Copilot, BBC, The Guardian etc.), about how Artificial Intelligence (AI) could soon replace developers.

I have been using Github Copilot, mainly for “enhanced” Visual Studio IntelliSense, code snippets, and refactoring. I have also found Github Copilot generated code samples and more context aware searching of documentation really useful. Overall, the experience has been positive with limited hallucinations and my productivity has got better as my “prompt engineer” skills improved.

Initially, I started building “hobbyist” AI applications for the edge with an Azure Percept and for customer projects use Ultralytics Yolo, Azure Machine Learning, Azure Cognitive Services, ML.Net, Open Neural Network Exchange (ONNX), Nvidia Jeston Modules, Seeedstudio reComputer Industrial J3011 – Fanless Edge AI Devices etc.

After some discussion with my co-workers I decided to explore how much Github Copilot helped building a “non-trivial” computer vision Proof of Concept (PoC). In previous blog posts I have written about capturing still images from a Security Camera so figured that was a good place to start.

Even though I have been using Copilot for a while I still say please.

using System;
using System.IO;
using System.Net.Http;
using System.Threading.Tasks;

class Program
{
    static async Task Main(string[] args)
    {
        string cameraUrl = "http://your-security-camera-url/image.jpg"; // Replace with your security camera URL
        string savePath = "C:\\path\\to\\save\\image.jpg"; // Replace with the path where you want to save the image

        using (HttpClient client = new HttpClient())
        {
            try
            {
                HttpResponseMessage response = await client.GetAsync(cameraUrl);
                response.EnsureSuccessStatusCode();

                byte[] imageBytes = await response.Content.ReadAsByteArrayAsync();
                await File.WriteAllBytesAsync(savePath, imageBytes);

                Console.WriteLine("Image retrieved and saved successfully.");
            }
            catch (Exception ex)
            {
                Console.WriteLine($"An error occurred: {ex.Message}");
            }
        }
    }
}

The generated code compiled first time, looked “ok” and worked with some of my security cameras when the cameraurl and save path were modified

I then modified the code so “sensitive” information was not stored in code or configuration files.

//using System;
//using System.IO;
//using System.Net.Http;
//using System.Reflection;
//using System.Threading.Tasks;

using Microsoft.Extensions.Configuration;

class Program
{
   static async Task Main(string[] args)
   {
      //string cameraUrl = "http://your-security-camera-url/image.jpg"; // Replace with your security camera URL
      //string savePath = "C:\\path\\to\\save\\image.jpg"; // Replace with the path where you want to save the image
      Model.ApplicationSettings applicationSettings;

   Console.WriteLine($"{DateTime.UtcNow:yy-MM-dd HH:mm:ss} 0.2.CameraHTTPClient starting");
#if RELEASE
         Console.WriteLine("RELEASE");
#else
      Console.WriteLine("DEBUG");
#endif
      // load the app settings into configuration
      var configuration = new ConfigurationBuilder()
           .AddJsonFile("appsettings.json", false, true)
      .AddUserSecrets<Program>()
      .Build();

      applicationSettings = configuration.GetSection("ApplicationSettings").Get<Model.ApplicationSettings>();

      using (HttpClient client = new HttpClient())
      {
         try
         {
            HttpResponseMessage response = await client.GetAsync(applicationSettings.CameraUrl);
            response.EnsureSuccessStatusCode();

            byte[] imageBytes = await response.Content.ReadAsByteArrayAsync();
            await File.WriteAllBytesAsync(applicationSettings.SavePath, imageBytes);

            Console.WriteLine("Image retrieved and saved successfully.");
         }
         catch (Exception ex)
         {
            Console.WriteLine($"An error occurred: {ex.Message}");
         }
      }
   }
}

namespace Model
{
   public class ApplicationSettings
   {
      public string CameraUrl { get; set; } = "";

      public string SavePath { get; set; } = "Images";
   }
}

Several of my security cameras (e.g. HILook T250H) required client to authentication so I added NetworkCredential my Copilot prompt.

// please write c# console application that uses httpclient and an httpget to retrieve image from security camera with NetworkCredential security
//
// Sure! Here's a C# console application that uses HttpClient and HttpGet to retrieve an image from a security camera with NetworkCredential security: 
// In this example:
//
// Replace http://your-security-camera-url/image.jpg with the actual URL of your security camera's image.
//
// Replace C:\\path\\to\\save\\image.jpg with the path where you want to save the retrieved image.
//
// Replace your-username and your-password with your actual credentials.
//
// This application uses HttpClient with NetworkCredential to authenticate the request, retrieves the image as a byte array, and saves it to the specified path. If there's an error during the process, it will be caught and displayed in the console.
//
// Happy coding! 🚀
//using System;
//using System.IO;
using System.Net;
//using System.Net.Http;
//using System.Threading.Tasks;

using Microsoft.Extensions.Configuration;

class Program
{
   static async Task Main(string[] args)
   {
      //string cameraUrl = "http://your-security-camera-url/image.jpg"; // Replace with your security camera URL
      //string savePath = "C:\\path\\to\\save\\image.jpg"; // Replace with the path where you want to save the image
      //string username = "your-username"; // Replace with your username
      //string password = "your-password"; // Replace with your password
      Model.ApplicationSettings applicationSettings;

      Console.WriteLine($"{DateTime.UtcNow:yy-MM-dd HH:mm:ss} 0.3.CameraHTTPClient starting");
#if RELEASE
         Console.WriteLine("RELEASE");
#else
      Console.WriteLine("DEBUG");
#endif
      // load the app settings into configuration
      var configuration = new ConfigurationBuilder()
           .AddJsonFile("appsettings.json", false, true)
      .AddUserSecrets<Program>()
      .Build();

      applicationSettings = configuration.GetSection("ApplicationSettings").Get<Model.ApplicationSettings>();

      using (HttpClientHandler handler = new HttpClientHandler { Credentials = new NetworkCredential(applicationSettings.Username, applicationSettings.Password) })
      using (HttpClient client = new HttpClient(handler))
      {
         try
         {
            HttpResponseMessage response = await client.GetAsync(applicationSettings.CameraUrl);
            response.EnsureSuccessStatusCode();

            byte[] imageBytes = await response.Content.ReadAsByteArrayAsync();
            await File.WriteAllBytesAsync(applicationSettings.SavePath, imageBytes);

            Console.WriteLine("Image retrieved and saved successfully.");
         }
         catch (Exception ex)
         {
            Console.WriteLine($"An error occurred: {ex.Message}");
         }
      }
   }
}

namespace Model
{
   public class ApplicationSettings
   {
      public string CameraUrl { get; set; } = "";

      public string SavePath { get; set; } = "Images";

      public string Username { get; set; } = "";

      public string Password { get; set; } = "";
   }
}

My Visual Studio 2022 solution with a project for each Copilot generated sample.

Summary

The Copilot generated code for my three “trivial” PoC applications compiled and worked with minimal modifications.

The Copilot generated code in this post is not suitable for production