in this project im using faster_rcnn_inception model with region proposal network (RPN)for generating region proposals.
the Task:
1-detect three types of fruit: apples, oranges and bananas
2-extract the fruit detected coordiants and send it to color_detection model and determine the color.
Data:
taken from Kaggle object detection datasets:
https://www.kaggle.com/mbkinaci/fruit-images-for-object-detection
requirements and packages :
in order to run this repostory please refer to the link below and download Tensorflow’s Object Detection API https://github.com/tensorflow/models
you may run the following commands to get the nesscary packages: pip install protobuf pip install pillow pip install lxml pip install Cython pip install jupyter pip install matplotlib pip install pandas pip install opencv-python pip install tensorflow
run the following commands from the models-master\research directory: python setup.py build python setup.py install
Testing the API: go to object_detection directory and enter the following command: jupyter notebook object_detection_tutorial.ipynb
after installing the nessarcy packages move everything in this repostory to models/research/object_detection remember to modify main.py for images paths that you wish to test the model on then you may run the file main.py by opening terminal and typing: python main.py
to include color_detection you will need to get my color_detection model from here : https://github.com/PalTAJ/knn-colorDetection
then get everything in it to models/research/object_detection directory. next open main.py (for the object detection) and enter uncomment the return_coordinate function call. next go to color_detection directory open main.py and add the following code (from link below) for single image color detection option : #in my case i only classifiy 3 objects. https://github.com/PalTAJ/rcnn-objectDetection/blob/master/changes_colord.py
then save it and run main.py by typing : python main.py
if you decided to train your own model please refer to tensorflow documentation :D