Laser & Optoelectronics Progress, Volume. 59, Issue 2, 0215002(2022)
Realizing Illumination Consistency in Augmented Reality Based on Shadow Detection
In the augmented reality field, it is a challenge to achieve the illumination consistency of virtual objects. To address the low shadow detection efficiency problem when virtual objects are endowed with shadow, a method based on the shadow area is proposed to construct a shadow volume to achieve the illumination consistency of virtual objects. The proposed method first performs superpixel merging based on the color distance similarity according to adjacent superpixel centers. Superpixel collection is obtained using an improved simple linear iterative clustering (Ⅰ-SLIC) algorithm on the images. The number of superpixel collections and the subsequent processing complexity are reduced accordingly. Then, a Gaussian mixture background model is employed to detect the shadow of the segmented image, and the shadow body is constructed using the shadow region and illumination parameters. Finally, the registration of the virtual object is completed according to the transformation matrix combined with the shadow volume for rendering. Experimental results demonstrate that the proposed method realizes the shadow rendering of virtual objects and greatly improves the realism of augmented reality applications. Compared with existing methods, the proposed method demonstrates an obvious advantage in terms of time efficiency.
Get Citation
Copy Citation Text
Guangyun Wu, Zhiping Zhou. Realizing Illumination Consistency in Augmented Reality Based on Shadow Detection[J]. Laser & Optoelectronics Progress, 2022, 59(2): 0215002
Category: Machine Vision
Received: Jan. 5, 2021
Accepted: Mar. 10, 2021
Published Online: Jan. 12, 2022
The Author Email: Wu Guangyun (616235675@qq.com)