I'm looking for freeware most preferably with open code that detects basic entities in pictures.
It means if I run it over a picture (gif, jpg, etc.) the system will find basic entities such as table, sky, man etc. and their location in picture.
Software Recommendations Stack Exchange is a question and answer site for people seeking specific software recommendations. It only takes a minute to sign up.Sign up to join this community
What you are asking for is Image Classification software and it is available as open source as a part of deep learning with image net. There is quite a lot of software to install so before you read any further you need to be aware that recognising portions of an image takes time - you will not be able to roll over a fresh image and have things recognised in it - instead you will need to run a classifier, preferably offline, and pre-classify your images. Also note that what you are asking for uses a lot of computer power, enough that most use it on an AWS Amazon EC2 g2.2xlarge image or better. The PyImageSearch blog has a series of tutorials on setting up and using such a setup.
pip install pillow
pip install scikit-learn
pip install h5py
pip install --upgrade --no-deps git+git://github.com/Theano/Theano.git
pip install keras
git clone https://github.com/fchollet/deep-learning-models
There is a tutorial on actually running this here or a number of books available.