Recognition of objects and their dimensions in Python
Budget: €30 – €250 EUR
Objective: Get the source code in Python so that from an image:
* identify and mark each object in the original image returning another image where the objects are boxed by strokes that form a cube,
* for each object provide an individual image only with the object previously framed by the cube,
* provide the actual dimensions (in the real world not in the image) of each object: height, width and depth (in cms.),
Details:
* in the image, among the different objects, there will be a reference object recognizable by the algorithm (this can have a certain color, for example it can be a red box) and whose dimensions are known and provided to the algorithm in order to serve as a reference to get the dimensions of the rest of the objects.
* all source code must be provided in Python so that it can be ran without problems in the Google Colab environment;
* the code must be provided and commented in detail;
* the freelancer must carry out several tests and take different photographs with objects of different volumes and distances in order to check the efficiency of the algorithm;
* it is allowed to install and call as many libraries as it is needed but they must be libraries of free code (example Open CV).
* identify and mark each object in the original image returning another image where the objects are boxed by strokes that form a cube,
* for each object provide an individual image only with the object previously framed by the cube,
* provide the actual dimensions (in the real world not in the image) of each object: height, width and depth (in cms.),
Details:
* in the image, among the different objects, there will be a reference object recognizable by the algorithm (this can have a certain color, for example it can be a red box) and whose dimensions are known and provided to the algorithm in order to serve as a reference to get the dimensions of the rest of the objects.
* all source code must be provided in Python so that it can be ran without problems in the Google Colab environment;
* the code must be provided and commented in detail;
* the freelancer must carry out several tests and take different photographs with objects of different volumes and distances in order to check the efficiency of the algorithm;
* it is allowed to install and call as many libraries as it is needed but they must be libraries of free code (example Open CV).