Palm Dates sorting machine
This project "Palm dates sorting machine" makes usage of computer vision to sort palm dates of different types. Using Intel NCS to inference the sorting process is very fast as inference on ncs is 150 ms and 350 ms on Movidius ...learn more
Project status: Under Development
Intel Technologies
OpenVINO,
Intel Opt ML/DL Framework,
Movidius NCS
Overview / Usage
This project "Palm dates sorting machine" makes usage of computer vision to sort palm dates of different types.
Our aim is to speed up sorting process as much as possible - decrease inference time - we use Intel Movidius NCS and openvino to convert tensorflow graph to run on intel MYRIAD VPU to accelerate the inference.
The machine have user interface to count dates and control speed of the belt.
Methodology / Approach
Using Tensorflow as a framework to train our data set and generate frozen graph - there is possibility to use caffe -
We train on data set of different palm dates types it consists of 1902 example of two types.
Then, we use openvino or NCSDK to convert frozen graph to graph can run on MYRIAD VPU in order to run the project using intel movidius ncsdk.
The machine have user interface to count dates and control speed of the belt.
Technologies Used
Computer vision
Tensorflow
retrain
object detection API
OpenCV
Openvino
NCSDK
Nanopi-M4
Collaborators
There are no people to show.