Poisson distribution model is the basis of data analysis for GM-APD lidar, but it is only applicable to the mirror reflected target under ideal conditions, and cannot accurately describe the photon triggering process of actual GM-APD lidar detection. For the actual target with rough surface, the negative binomial distribution with M as the parameter can describe the photon distribution model more accurately. In order to solve this problem, this paper compares and analyzes Poisson distribution triggering model and the negative binomial distribution triggering model that conforms to the triggering situation of the echo signal of the target with rough surface, considering the differences in the triggering probability under different noise and signal levels. The results show that the trigger probability curve corresponding to the trigger model based on negative binomial distribution is lower in peak value, wider in bottom value and fatter overall than that of the trigger model based on Poisson distribution, and the difference between the two is more prominent under the conditions of low noise level and high signal level. This study has guiding significance for the signal extraction research based on different surface echoes.
In the military and civil fields, detecting small aircraft is of great significance. In recent years, the rapid development of LiDAR technology has made it possible to detect small aircraft at long distances. However, the scale change and attitude change make the detection difficult. Therefore, a detection network of multi-attitude small aircraft based on LiDAR anchorfree is proposed in this paper. The network structure is improved on the basis of the CenterNet network; using the encoder-decoder network structure, the extended convolutional module is designed to improve the receptive field and obtain the multi-scale information of the object. The IOU sensing branch is added to the detection header of the network to improve the localization accuracy of the object. The experimental results show that the detection accuracy of the improved network on the self-built simulation data set is 2.12% higher than that before the improvement and finally reaches 92.45%. Therefore, using this method can effectively improve the detection accuracy of the object.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.