Get started

We assist civil and structural engineers, inspectors, and organizations in continuous monitoring and realtime assessment of the conditions of road bridges to enhance their infrastructure maintenance program. Our AI algorithms process images or video frames to detect various types of bridge defects, such as corrosion, cracks, abnormal spacing, functional disorder of bearings, spalling/exposed rebar, etc. and identify the severity ratings of those defects.

Our current AI-based solution is optimized for concrete bridges.

We have developed a customizable API that can be readily used for any use-case at hand or can be fine-tuned further to expand the list of bridge defects. The flexibility of our API allows developers easily modify and seamlessly integrate the end-product into their own workflows.

You will need a key to use our API . please reach out to us at [email protected]

Try It Now

Examples:

Disclaimer: We have set a filter for image analysis. If an image is caught by the filter, the expected result might not display correctly.
However, direct API usage will work for all elements, regardless of the filter.

Drag image file here or

Browse from your computer

Class Labels JSON Data

x_min	y_min	x_max	y_max

Box Class	Box Confidence

Installation

To install the API client, you can use pip (Python's package manager). Ensure you have Python installed on your system. Run the following command in your terminal or command prompt:

                
                    pip install smretrofit
                
            

After installation, you can import the API client into your Python scripts and start making requests. Ensure that you have the correct authentication credentials (API key and password) to interact with the API successfully.

For more information and documentation, visit the GitHub repository.

If you encounter any issues or have questions, please check the GitHub Issues page for support.

Parameters

To detect construction defects and ratings, make a POST request to the following URL:
https://api.somikoron.ai/api

CLASS PARAMETERS

Following Paramiters are for API Class

IMAGE FUNCTION PARAMETERS

Following Paramiters are for image_data function

VIDEO FUNCTION PARAMETERS

Following Paramiters are for video_data function

VIDEO SAMPLE FUNCTION PARAMETERS

Following Paramiters are for video_data function

Code Example

A basic code example is provided below to help you get started with using the API. This example demonstrates how to make a request to the API and process the results. You can customize it based on your specific needs.

import Retrofit as api

url = "https://api.somikoron.ai/api/"

auth_key="your_api_key"

auth_pass="your_api_password"

api = api.Retrofit(url, auth_key, auth_pass)

if __name__ == "__main__":

    file_path = "your_file"

    data = api.image_data(file_path, save=True)

Returned Data Example

The API will return a JSON object containing the detected class names and the result data for each processed frame. If the save parameter is set to true, the response will also include the rendered image data in base64 format.

                
                    [

                         {'cls': 
 
                            {'0': 'defect_class_name', '1': 'defect_class_name', '2': 'defect_class_name', '3': 'defect_class_name', ...}, 
                    
      'data': [ 
 
                                {'box_xyxy': [x, y, x, y], 
                    
              'box_cls': 2, 

                                 'box_conf': 0.851271390914917} 

                                 ] 

                         }, 

                    

                         {'cls': 
 
                            {'0': 'rating_class_name', '1': 'rating_class_name', '2': 'rating_class_name', '3': 'rating_class_name', ...}, 
                    
      'data': [ 
 
                                {'box_xyxy': [x, y, x, y], 
                    
              'box_cls': 2, 

                                 'box_conf': 0.851271390914917} 

                                 ] 

                         }, 


                    ]    
                
            

As for the video, the JSON data will be structured as a list, where each item in the list represents a frame's data, ordered serially by frame.

Errors

The Construction Defect and Rating Detection API uses the following error codes: