Build an Image classification model for flower species

• The purpose of this notebook is to build a Image classifier which would be capable of predicting the species of flowers
• We compare and contrast the performance of various machine learning algorithms VS Deep learning algorithms in an effort to demostrate how traditional ML algorithms perform poorly on Image Data
• Details of the problem statement , data set , summary of the code/solution , sample output/Prediction from the program and final result of the project are listed in the sections to follow.

Computer Vision can be used to classify images .In this partcular use case , a university conducting research involving the understanding of the characteristics of flowers,require an automation which can create a classifier capable of determining a flower’s species from a photo

The dataset comprises of images from 17 plant species. It can be downloaded from TensorFlow [tflearn.datasets.oxflower17 as oxflower17 ]

Botanical research

The code aims at building a image classfifier

• We begin by doing an Exploratory Data analyses and Visualisation/viewing of the images
• We then do the required pre-processing for the data to make it compatible with the various models to be built.Each model would require the data to be prepared specifically for that model to ensure compatibility.
• Essentially , the approach used in this worksheet is to try out multiple algorithms to build the classifier and choose the best contender
• Multiple 'traditional' ML algorithms like LR,KNN,SVM,Random Forest,Ada Boost, Gradient Boostare tried and results were recorded
• We then try deep learning algorithms like ANN and a CNN(build custom ANN & CNN networks) and try to solve the problem at hand .Results are recorded
• Finally we try two pretrained CNN with a base of VGG16 and RESNET50 respectively and observe that CNN with base VGG16 gives us the best results
• Pre-trained CNN(on ImageNet) with VGG16 Network as a base performed the best giving us 87% on Test Data
• Refer python worksheet Project_P4_ImageClassfication_BotanicalResearch_Flowers.ipynb for the solution

Here is a sample result/output from the program/model

• PGP Course Material
• Evaluator Feedback