Classifying Oranges With Orangelib:
Classification problem is one of the major challenges in machine learning. Computer vision is an important aspect of classification problem in which objects recognized by a computer can be classified into specific categories.
Implementing computer vision can be challenging, such as training a model to perform accurate prediction on an object. Training a model to accurately recognize a specific object can be tasking and it becomes more tasking to implement a computer vision model that will be able to classify the difference between one object and another.
Orangelib:
Orangelib is a simple library created to facilitate easy implementation of computer vision in real problems. It is a library for classifying oranges into two categories: ripe and unripe.
This classification can be achieved with just few lines of code.
We shall look into the lines of code below.
The code above import in the class orangeclassifier from orangelib for performing prediction on images.
The trained model for classifying oranges can be downloaded as a release from github
The path to image that will be predicted.
We print out the fruit name and level of prediction confidence
Install orangelib using:
- pip install orangelib
Orangelib will be used to perform inference on four images.
Sample1:
Output1:
Sample2:
Output2:
Sample3:
Output3:
Sample4:
Output4:
We can clearly observe that the predictions on the four images are quite accurate, 99% accurately predicting both ripe and unripe oranges.
- Install orangelib and test it with as many samples as you wish.
- Visit the github repository hosting orangelib.
