Machine Learning for Classifying Images with Motion Blur

Motion blur in an image results from movements of objects within a scene or from the imaging system itself. In applications such as high-speed license plate recognition, motion blur introduces image artifacts leading to challenges in image classification. In this work, we investigate machine learning techniques for classifying images with motion blur using convolutional neural networks. In particular, we explore how different motion directions and lengths affect the predictive performance of our classification model. We used the MNIST dataset, which contains 70,000 images of handwritten digits, to generate training and testing datasets of images with motion blur using MATLAB. Specifically, we considered motion blurs at various angles and lengths to analyze the effects of different motion blurs on the classification accuracy of images of digits. We found that our model shows very high accuracy when training and testing on datasets with the same type of motion blur. In contrast, training and testing on datasets with different motion blurs result in lower accuracy. We describe how we can improve overall classification performance and offer insights on what additional information can be inferred from the MNIST dataset with motion blur.

I. Introduction

Motion blur is a common occurrence in scenarios where there is relative motion between the camera and targets during exposure. Most machine learning approaches are trained and tested on high quality image datasets, thus images affected by motion blur are sure to impact the classification accuracy of a given machine learning approach. In settings where we might be interested in identifying of images affected by blur, it is common to first deblur or enhance the image quality before classifying an image.