Skip to content

Circuit-Digest/Parking-Detection-Using-ESP32-Cam

BranchesTags

Repository files navigation

🅿️ Parking Space Detection Using ESP32-CAM with CircuitDigest Cloud API

A smart, AI-powered parking space detection system using the ESP32-CAM and CircuitDigest Cloud API. Captures a real-time image of a parking area and instantly detects the number of occupied and free spaces — wirelessly, affordably, and without any ML training!

📌 Table of Contents

🌍 Overview

Imagine driving into a crowded parking area and not knowing whether a space is available or not. This project solves that problem by creating a smart and automated parking space detection system using the ESP32-CAM and CircuitDigest Cloud.

With just a button press, the device:

  1. Captures a real-time image of the parking area
  2. Sends it wirelessly over the internet to the cloud AI
  3. Receives back a breakdown of occupied vs. free parking spaces
  4. Displays the result in the Serial Monitor within seconds

Low-cost. Wireless. Real-time. No ML training required. ⚡

⚙️ How It Works

[ Push Button Pressed ]
        ↓
[ ESP32-CAM Captures Image of Parking Area ]
        ↓
[ Image Sent via Wi-Fi over HTTPS ]
        ↓
[ CircuitDigest Cloud API Processes Image with AI ]
        ↓
[ Result Returned: Occupied Slots + Free Slots + Confidence Values ]
        ↓
[ Output Displayed on Serial Monitor ]

Tip: For testing, a real-time image is not required — any parking lot image from the web works too! Ensure sufficient lighting for better detection accuracy.

🧰 Components Required

S.No Component Purpose
1 ESP32-CAM Microcontroller with built-in camera & Wi-Fi
2 Push Button Triggers image capture on press
3 Breadboard Simplifies and organizes circuit connections
4 USB-to-Serial (FTDI) Adapter (if needed) For programming standard ESP32-CAM without onboard USB
5 USB Cable Powers the system via laptop/PC

⚠️ Note: If you are using the standard ESP32-CAM (without onboard USB), you need a USB-to-Serial (FTDI) adapter for programming:

  • FTDI TX → ESP32-CAM RX (U0R)
  • FTDI RX → ESP32-CAM TX (U0T)
  • GND → GND
  • Hold GPIO0 LOW during upload to enter flash mode.

🔌 Circuit Diagram

The push button is connected to GPIO13 of the ESP32-CAM to trigger image capture.

ESP32-CAM            Push Button
---------            -----------
 GPIO13  ───────────  One Terminal
  GND    ───────────  Other Terminal

Connect the ESP32-CAM to your laptop via USB for power. Refer to the circuit diagram image in the repository for a detailed visual reference.

🚀 Getting Started

Step 1: Create a CircuitDigest Cloud Account

Go to the CircuitDigest Cloud website, create a free account, and log in. On the homepage, scroll down and click on the Parking Space Detection feature.

Step 2: Get Your API Key

  • Inside the Parking Space Detection page, navigate to "Get Started" on the left panel.
  • Your API Key will be displayed there — copy and save it.
  • Adjust the confidence level based on your accuracy requirements.

📊 API Limits: 15 requests/day and 100 requests/month on the free tier.

Step 3: Test the API Virtually

  • Use the "Try API" feature on the dashboard.
  • Upload an image of a parking area with a mix of occupied and empty slots.
  • Click "Run Test" — within seconds, the system displays:
    • Number of filled slots
    • Number of empty slots
  • Test with different images to evaluate detection accuracy before deploying hardware.

⚠️ Note: Each "Try API" test counts toward your daily/monthly API usage limit.

Step 4: Hardware Setup & Code Upload

  1. Connect all components as per the circuit diagram.
  2. Open Arduino IDE and install the ESP32 board package.
  3. Open the project sketch and fill in your credentials:
const char* WIFI_SSID  = "Your_WiFi_SSID";
const char* WIFI_PASS  = "Your_WiFi_Password";
const char* API_KEY    = "Your_API_Key_Here";
  1. Select AI Thinker ESP32-CAM as the board in Arduino IDE.
  2. Upload the code (hold GPIO0 LOW if using FTDI adapter).
  3. Open Serial Monitor at 115200 baud.
  4. Point the camera at a parking area and press the push button.
  5. The parking status result appears on the Serial Monitor within seconds!

💻 Code Explanation

1. Library Includes

#include "esp_camera.h"
#include <WiFi.h>
#include <WiFiClientSecure.h>

Includes libraries for ESP32 camera control, Wi-Fi connectivity, and secure HTTPS communication. The secure client enables encrypted data transfer between the ESP32-CAM and the cloud API.

2. Wi-Fi & API Configuration

const char* WIFI_SSID  = "YourSSID";
const char* WIFI_PASS  = "YourPassword";
const char* API_KEY    = "YourAPIKey";
const char* serverName = "www.circuitdigest.cloud";
const char* serverPath = "/api/v1/parking-detection/detect";

Defines Wi-Fi credentials and the cloud API server details. The serverPath points to the parking space detection model. The API key is used for authentication when sending image data to the cloud.

3. Camera Pin Mapping

#define PWDN_GPIO_NUM  32
#define XCLK_GPIO_NUM   0
#define Y2_GPIO_NUM     5
#define PCLK_GPIO_NUM  22
// ... (remaining pins)

Maps ESP32 GPIO pins to the camera module's data lines, clock, and synchronization signals. A correct pin configuration is essential for stable image capture and system reliability.

4. Camera Initialization

void initCamera() {
  camera_config_t cfg = {};
  cfg.pixel_format = PIXFORMAT_JPEG;
  cfg.frame_size   = FRAMESIZE_VGA;
  esp_camera_init(&cfg);
}

Initializes the camera in JPEG format at VGA resolution for an optimal balance of transmission speed and image quality. If initialization fails, the system halts to prevent further errors.

5. Capture & Send to API

camera_fb_t* fb = esp_camera_fb_get();   // Capture image
String result = sendImageToAPI(fb);       // Send to cloud API
esp_camera_fb_return(fb);                 // Free frame buffer

On button press, captures a JPEG image and sends it to the CircuitDigest Cloud API via HTTPS POST. The frame buffer is released after transmission to free memory. The API processes the image using its AI model and returns the parking detection result.

📊 Output

After pressing the push button, the Serial Monitor displays:

Connecting to WiFi...
WiFi Connected!
Button Pressed - Capturing Image...
Image Captured. Sending to API...

--- Parking Space Detection Result ---
Total Spaces Detected: 10
  Occupied Slots : 7  (Confidence: 96.3%)
  Empty Slots    : 3  (Confidence: 94.1%)

Parking Status: LIMITED SPACES AVAILABLE

The result includes occupied count, empty count, and the confidence score for each category.

✅ Advantages & Limitations

S.No Advantages Limitations
1 Real-time parking detection within seconds Cannot work without cloud API access
2 Low-cost system using ESP32-CAM with built-in camera & Wi-Fi Requires an active internet connection
3 No expensive hardware or powerful processors needed Blurry images may produce incorrect results
4 Fully wireless communication without extra modules Limited by daily/monthly API usage limits
5 Small size and portable design Captures single images, not continuous live video

Why CircuitDigest Cloud over alternatives like Edge Impulse? Edge Impulse requires dataset collection, model training, optimization, and deployment — time-consuming and complex. CircuitDigest Cloud provides a ready-to-use API that eliminates all of that, reducing development time and allowing focus on system integration rather than AI model design.

🛠️ Troubleshooting

🔇 Issue 1: No Output in Serial Monitor

  • Cause: Incorrect COM port or baud rate setting.
  • Fix: Set baud rate to 115200, verify the correct COM port in Arduino IDE, and confirm the USB cable and board are properly connected.

🔄 Issue 2: ESP32-CAM Restarting Continuously

  • Cause: Insufficient or unstable power supply.
  • Fix: Use a proper 5V regulated external power source. Avoid relying only on weak USB connections and check for loose wiring.

🚫 Issue 3: API Usage Limit Exceeded

  • Cause: Too many test requests consumed the daily/monthly quota.
  • Fix: Check your usage in the CircuitDigest Cloud dashboard. Wait for the limit to reset or upgrade your plan.

📷 Issue 4: Image Capture Failure

  • Cause: Loose camera connections or memory limitations.
  • Fix: Ensure the camera module is securely connected. Try reducing image resolution (e.g., VGA) to avoid memory-related issues.

🌐 Issue 5: API Connection Error or Timeout

  • Cause: Unstable internet connection, invalid API key, or incorrect server configuration.
  • Fix: Verify your internet connection, double-check the API key, and ensure HTTPS is properly configured in the code.

🔢 Issue 6: Incorrect Detection Results

  • Cause: Poor image quality, bad lighting, or low confidence threshold.
  • Fix: Use clear images with proper lighting. Adjust the confidence level in the CircuitDigest Cloud dashboard for better accuracy.

❓ FAQ

1. Why is ESP32-CAM used in this project?

ESP32-CAM has a built-in camera module and Wi-Fi capability, making it ideal for capturing and transmitting images over the internet. It is cost-effective and perfect for IoT-based image processing applications.

2. Why is the cloud API used instead of local processing?

The ESP32-CAM has limited memory and processing power, which makes running complex AI models locally impractical. The cloud API performs heavy inference on powerful servers, ensuring better accuracy and real-time performance.

3. What happens when the push button is pressed?

The ESP32-CAM captures an image of the parking area and sends it to the CircuitDigest Cloud API. The API processes the image and returns the number of occupied and empty parking slots, which are then displayed on the Serial Monitor.

4. How accurate is the parking detection system?

Accuracy depends on image quality, lighting conditions, and camera positioning. Clear images with good lighting and fully visible parking slots significantly improve detection accuracy.

5. Can this system work without the internet?

No. The system requires an active internet connection because all image processing is performed on the cloud server. Without internet, the ESP32-CAM cannot send data to the API.

6. What are the limitations of this system?

The system depends on internet connectivity and API usage limits. It may also produce incorrect results if image quality is poor or parking slots are not clearly visible in the captured image.

🔗 Relevant Links

Made with ❤️ by CircuitDigest | Powered by CircuitDigest Cloud AI

About

No description, website, or topics provided.

Resources

Readme
Activity
Custom properties

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors