Our proposed idea will make use of an object detection model to scan the provided image for the item in question. If the model is not able to detect the object outright, it will not let the user post his item until he provides a valid photo. If the photo contains clutter and other irrelevant details, the model will scan the photo, locate the object in question and then crop the image to only contain the object. This will lead to less misleading or irrelevant features, and will automate the process instead of requiring manual inspection. In addition, this process will be much faster in taking action than manual inspection.

node server.js 

http://localhost:8080/

1- First the server starts listening to a specific port.

Server started in 8080

2- Second connecting to mongodb is successful.

Conntected To Mongo Database

3- Finally the model (Coco-Ssd) is loaded successfully, this may take several seconds before appearing.

model uploaded successfully