Vehicle Logo Recognition Based On Convolutional Neural Network And K-Nearest Neighbor

Vehicle logo recogniton is an improtant research field in computer vision and plays an important role in the field of intelligent transportation system. Vehicle logo recognition technology is a practical technology for motor vehicle brand informationthrough image processing and automatic identification method based on a digital image or video signal stream as the object. The vehicle recognition system includes includes two key technologies of vehicle logo location and vehicle logo recognition.Due to the differences of vehicle logo, the different environmental conditions anduncertainty of vehicle logo location.A good vehicle logo location and recognition method is full of challenging.In this paper, we put forward a coarse-to-finemethod for logo location In the vehicle logo recognition, we use convolutional neural network(CNN) to train logo features and classify the logo, the experiment obtained high recognition rate. Then we combined with the traditional k nearest neighbor algorithm with theconvolutional neural network extracting features, to improve the vehicle logo recognition, reached 96.4% correct recognition rate.Finally,we put forward a rapid and efficient a method for vehicle logo detection and recognition based on Selective Search and SPP-net。...