Objective: The objective of this study is to develop a novel algorithm on a mobile system that can warn drivers about the possibility of a collision with a pedestrian. The constraints of the algorithm are near-real-time detection speed and a good detection rate.
Method: Histogram of gradients (HOG)-based detection is widely used in pedestrian safety applications; however, it has low detection speed for real-time systems. Hence, it has no direct usage for mobile systems. In order to achieve near-real-time detection speed, partial Haar transform predetections are applied to an image before HOG detection. The partial and HOG detections are merged and a score-based confidence level is defined for the final detection phase. In this way, the outcome is prioritized and different warning levels can be issued to warn the driver before a possible pedestrian collision.
Results: The proposed algorithm provides an increase in detection speed (from 46 to 76 fps) and detection rate (from 80 to 91%) with respect to HOG-based pedestrian detection. It also improves confidence of the results by multidetection merging and score assignment to detections.
Conclusions: Performance improvement of the algorithm is compared with respect to state-of-the-art detectors/algorithms. Based on the detection rate and detection speed performance, it can be concluded that the proposed algorithm is suitable to be used for mobile systems to warn drivers about the possibility of collision with a pedestrian. 相似文献
The intensity of available light is important in determining how well a diurnal animal can distinguish color. Primates with different types of color vision may exhibit behaviors that maximize visual contrast during critical activities such as feeding. We hypothesized that (1) trichromatic taxa will feed in a wide range of light conditions because color constancy permits stabilized color appearance across changes in illumination, and (2) that taxa with a high proportion of dichromatic individuals will tend to feed at higher light levels to increase color contrast. We recorded light levels during feeding bouts of seven primate taxa with varying degrees of color vision: the dichromatic Lemur catta, two polymorphic species, Propithecus v. verreauxi and Ateles geoffroyi, and the routine trichromats, Alouatta palliata, Colobus guereza, Piliocolobus badius, and Cercopithecus ascanius. Results were equivocal for our hypotheses. While routinely trichromatic taxa used varying light levels, the pattern of results did not differ from the dichromatic Lemur catta. However, polymorphic taxa not only sought the highest light, but females, which are the only individuals in polymorphic taxa that can be trichromatic, fed in higher light levels than males when eating non-green foods. This result is consistent with selection operating to maintain a balanced polymorphism in these taxa, though the connection between light levels and color vision type for the females is unclear. Our results further suggest that trichromatic vision may afford a selective advantage because it permits foraging under a greater range of light levels. 相似文献
Among the most familiar sexual signals are red, yellow, and orange sexual traits pigmented by carotenoids. Many birds can
detect near-ultraviolet (UV) light, and UV signals can play key roles in mate choice. Grouse (Tetraonidae) exhibit bright
carotenoid-dependent sexual ornaments, their supra-orbital combs, which to humans appear orange-red. Combs also reflect in
the UV, which is not visible to humans but is likely to be visible to grouse. In male red grouse Lagopus lagopus scoticus, we show that comb UV reflectance decreases with increasing comb size and redness. By removing the epidermis of combs, where
carotenoid pigments are, we show that the UV reflectance is a property of the dermis, underneath the red pigmented epidermis.
Carotenoid pigmentation of combs acted as a mask to reduce reflectance by the dermis in the range 400–550 nm and in the UV,
300–400 nm. Patagium skin (non-ornamental skin under the wing) also reflects in the UV, but epidermis removal on this bare
part tended to reduce UV reflectance, whereas removal of the red epidermis of combs increased UV reflectance. Males in better
condition (greater body mass relative to size) had bigger and redder combs, but with less UV. Thus, carotenoid pigments of
grouse combs are deposited on a white background with significant UV reflectance, which can influence how the signal is perceived
by conspecifics. Carotenoid-based traits exhibit UV reflectance in a number of species, but how UV reflectance and carotenoid
pigmentation influence colour remains little known for integumentary ornaments compared to plumage traits. UV vision is not
uncommon in birds and other animals, so future studies should investigate how UV reflectance influences the perception of
carotenoid-based signals of quality. 相似文献
Objective: Fatal vision goggles (FVGs) are image-distorting equipment used within driver education programs to simulate alcohol-related impairment. However, there is no empirical evidence comparing the behavioral effects associated with wearing FVGs to alcohol intoxication. The purpose of this study was to determine the validity of FVGs in producing alcohol-related impairment in simulated driving.
Methods: Twenty-two healthy males (age: 23 ± 3 years, mean ± SD) participated in a placebo-controlled crossover design study involving 4 experimental trials. In each trial, participants completed a baseline level simulated driving task followed by an experimental driving task, involving one of 4 treatments: (1) a dose of alcohol designed to elicit 0.080% breath alcohol concentration (BrAC; AB), (2) an alcohol placebo beverage (PB), (3) FVG (estimated % blood alcohol concentration [BAC] 0.070–0.100+), and (4) placebo goggles (PGs). The driving tasks included 3 separate scenarios lasting ~5 min each; these were a simple driving scenario, a complex driving scenario, and a hazard perception driving scenario. Selected lateral control parameters (standard deviation of lane position [SDLP]; total number of lane crossings [LCs]) and longitudinal control parameters (average speed; standard deviation of speed [SDSP]; distance headway; minimum distance headway) were monitored during the simple and complex driving scenarios. Latency to 2 different stimuli (choice reaction time [CRT]) was tested in the hazard perception driving scenario. Subjective ratings of mood and attitudes toward driving were also provided during each of the trials.
Results: Neither placebo treatment influenced simulated driving performance. Mean BrAC was 0.060 ± 0.010% at the time of driving on the AB trial. Lateral control: In the simple driving scenario, SDLP and LC were not affected under any of the experimental treatments. However, in the complex driving scenario, significantly greater SDLP was observed on both the FVG and AB trials compared to their respective baseline drives. LC increased significantly from baseline on the AB trial only. Longitudinal control: Speed was not affected by any of the experimental treatments; however, SDSP increased significantly from baseline on the FVG trial. A significant reduction in distance headway and minimum distance headway was detected on the FVG trial compared to baseline. Hazard perception: Neither AB nor FVG trials were influential on CRT. Subjective mood ratings were significantly altered on the AB and FVG trials compared to baseline and placebo conditions. Participants reported reduced willingness and ability to drive under the active treatments (AB and FVG) than the placebo treatments (PB and PG).
Conclusions: FVGs may have some utility in replicating alcohol-related impairment on specific driving performance measurements. Hence, the equipment may offer an alternative approach to researching the impact of alcohol intoxication on simulated driving performance among populations where the provision of alcohol would otherwise be unethical (e.g., prelicensed drivers). 相似文献