红外技术, 2017, 39 (1): 53, 网络出版: 2017-03-29
一种三角函数权重的图像拼接算法
An Image Fusion Algorithm Based on Trigonometric Functions
加权平均融合 渐入渐出融合 三角函数 权重 图像融合 曝光差异 weighted average image fusion gradual image fusion trigonometric functions weight images fusion different exposure index
摘要
针对传统加权平均融合算法和渐入渐出融合算法仍然存在相对明显的拼接痕迹, 提出一种三角函数权重的图像拼接算法。首先, 对参考图计算图像重叠区域从左边界开始每一列像素所占重叠区的比例, 将其用相应的角度表示; 然后计算角度对应余弦值的平方, 将此结果作为参考图的权重信息; 对于目标图, 计算靠近右边界的每一列所占重叠区的比例, 并用角度表示, 然后计算该角度对应正弦值的平方, 将此结果作为目标图的权重; 最后用计算得到的 2个非线性的权重对两幅图进行图像拼接。实验结果表明无论摄像机是否在曝光差异较大情况下进行拍摄, 改进的图像融合算法效果更好。
Abstract
Many images stitching algorithms such as the weighted average image fusion and the gradual image fusion, could not make the stitching gap disappear. A new image fusion method based on trigonometric functions was proposed. Firstly, the overlap area ratio was calculated for each column of pixels, and was represented with corresponding angles. Then, the square of the cosine value was calculated, as weighted information of reference map. The square of the sine value was calculated as weighted information of target image. Finally, the two images are fused by using those two nonlinear weights. The experimental results show that the improved image fusion algorithm has better effects, whether or not the camera is operating in large exposure differences.
汪丹, 刘辉, 李可, 周威. 一种三角函数权重的图像拼接算法[J]. 红外技术, 2017, 39(1): 53. WANG Dan, LIU Hui, LI Ke, ZHOU Wei. An Image Fusion Algorithm Based on Trigonometric Functions[J]. Infrared Technology, 2017, 39(1): 53.