SiftKeyPre: A Vehicle Recognition Method Based on SIFT Key-Points Preference in Car-Face Image
Vehicle recognition from images produced in roads bayonet provides important clues to solve vehicle crime cases. Its accuracy is not enough to meet the requirement in real conditions. We proposed a vehicle recognition method, SiftKeyPre, based on SIFT(Sca
- PDF / 419,606 Bytes
- 15 Pages / 439.37 x 666.14 pts Page_size
- 50 Downloads / 158 Views
Laboratory of Parallel Software and Computional Science, Institute of Software, Chinese Academy of Sciences, Beijing 100190, China [email protected] 2 School of Infomatics Science and Technology, Shijiazhuang Tiedao University, Shijiazhuang 050043, China 3 Institute of Applied Mathematics, Hebei Academy of Sciences, Shijiazhuang 050081, China
Abstract. Vehicle recognition from images produced in roads bayonet provides important clues to solve vehicle crime cases. Its accuracy is not enough to meet the requirement in real conditions. We proposed a vehicle recognition method, SiftKeyPre, based on SIFT(Scale-invariant feature transform) key points preference for car-face images. Firstly, SiftKeyPre choices the SIFT key points following the DualMax algorithm to get a DualMax set. Meanwhile, Lowe set is defined as another one following Lowe algorithm. Secondly, we define a DL set under an intersection operation on DualMax set and Lowe set. For positive examples training images, we count the appearance times of each key point of DL set to compute the attention degree of each key point in base image. Finally, matching degree between the base image and a target image is evaluated with the attention degree of each matched points. SiftKeyPre method confirms a testing image based on its matching degree. Experiments results show that, under a given recall constraints, the precision of SiftKeyPre method is better than FLANN and Lowe. SiftKeyPre’s computational complexity is closed to that of Lowe. Comparing with other algorithms based on training, SiftKeyPre is of lower training intensity. Keywords: Car-face image · SIFT key points · Preference · Attention degree · Matching degree · Recognition
1
Introduction
In modern societies, the insecurity and threat events are increasing. Vehicle recognition from high-definition vehicle images produced in roads bayonet is an important source of clues on which public security departments solve cases of vehicle crime relies. For the phenomenon of faking and sheltering vehicle license plate, vehicle recognition system can not recognize car types correctly just based on the license plate. In a medium-sized city, more than 10000 images are captured at each major road bayonet by high-definition cameras per hour. At present, policeman selects out certain type of vehicle, for example black PASSAT, by “eyes of human”. © Springer-Verlag Berlin Heidelberg 2015 H. Zha et al. (Eds.): CCCV 2015, Part II, CCIS 547, pp. 344–358, 2015. DOI: 10.1007/978-3-662-48570-5_34
SiftKeyPre: A Vehicle Recognition Method Based on SIFT Key-Points Preference
345
This retrieval process is non-efficiency and tedious. It is urgency for investigation on automatic recognition algorithms to identify suspect vehicle. The basic problem in computer vision research is object classification and detection. The image object recognition is an important branch with more than fifty years history [1]. Image object recognition algorithms are divided into two basic categories, algorithms based on low-level feature and deep learn
Data Loading...
