In order to solve lane line detection in difficult traffic conditions, such as shadow occlusion, signpost degradation, curves, and tunnels, numerous models have been proposed. However, most of the existing models conduct independent single-frame image detection, which makes it difficult to utilize the continuity of driving images and is ineffective in challenging scenes. To this end, we suggest a spatiotemporal information processing model for lane line recognition that enhances critical features. In order to properly learn the correlation between continuous images, we first employ a convolutional gated recurrent unit to process spatiotemporal driving information on the basis of U-Net. Second, the pyramid split attention (PSA) module is used to enhance or suppress the obtained feature expressions. Finally, the skip connection is used to fuse the features of different scales encoded by each stage with the features processed by PSA and gradually restore to the original image size. Experiments on the TuSimple dataset demonstrate that our model outperforms representative lane line detection networks in challenging driving scenes, with an F1-measure of up to 94.302%.
The object distance of the high-resolution satellite camera will be changed when the camera is scroll imaging, which will
cause not only the alteration of the image scale, but also the variation of the velocity-height ratio (V/H) of the satellite.
The change of the V/H of the camera will induce the asynchronization between the image motion and the traveling of the
charge packet on the focal plane, which will deteriorate the image quality of camera seriously. Thus, the variable
regulation of the relative velocity and the height of the scroll imaging of the satellite were researched, and the expression
the V/H was deduced. Based on this, the influence of the V/H on the image quality was studied from two variable
factors: the latitude and the scroll angle. To illustrate this effect quantitatively, using a given round polar orbit, the
deterioration of the image quality caused by the scroll imaging was calculated for different integral number of the
camera, and regulation interval of the row integration time and the range of the scroll angle were computed. The results
showed that, when the integral number of the camera is equal to 32 and 64, the permitted scroll angle are equal to 29.5°
and 16° respectively for MTFimage motion >0.95, which will give some helpful engineering reference to learn how the
image quality changes during scroll imaging of the satellite camera.
To measure sub-pixel image motion of sequential images which is captured at high frame rate, the joint transform
correlator (JTC) is used. The relative image motion of two adjacent images can be measured by inputting these images
into JTC. The principle of this method is described, photo-electrical devices are selected, and an experimental platform is
built. Based on which, the measurement performances of JTC are researched, including the influence of scene structure
on measurement accuracy, effect of the size of input image on measurement accuracy, and the measurement range of the
image motion of JTC. After doing these, the over-all properties of JTC is verified by using a sample which containes 50
different random image motions. The results show that, the JTC can measure sub-pixel image motion of two adjacent
images entirely, and the accuracy is not variable with the contents of input images. The measurement error submits to
normal distribution, which means zero and RMS is no more than 0.12 pixels under conspicuous level equal to 0.05.
Lastly, the source of error which deteriorates measurement accuracy is analyzed simply.
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.