基于机器视觉的二维动画图像分割方法

Segmentation method of two-dimensional animation image based on machine vision

为保证动画的制作效果,需较好地完成二维动画图像的处理,因此,提出基于机器视觉的二维动画图像分割方法。采用生成对抗网络模型处理二维动画图像中的雾状现象,同时,采用多尺度Retinex算法补偿去雾后二维动画图像的色彩,在保证图像的色彩效果后,将色彩补偿后的二维动画图像输入改进的二维直方图方法中,完成二维动画图像分割。测试结果表明,该方法可极大程度去除图像中的雾状情况,提升图像的清晰度以及图像的色彩效果;能够精准完成二维动画图像中目标图像分割处理,获取完整的目标图像;分割后图像的一致性测度和区域间对比度结果均在0.933以上,分割后图像的均匀性和对比度较好;形状测度结果均在0.927以上,分割后图像质量较好,满足不同背景复杂程度的二维动画图像分割需求。

In order to ensure the production effect of animation,it is necessary to complete the processing of two-dimensional animation image.Therefore,this paper proposes a two-dimensional animation image segmentation method based on machine vision.This method uses the generation of confrontation network model to deal with the fog phenomenon in the two-dimensional animation image.At the same time,the multi-scale Retinex algorithm is used to compensate the color of the two-dimensional animation image after fog removal.In order to ensure the color effect of the image,the two-dimensional animation image after color compensation is input into the improved two-dimensional histogram method to complete the two-dimensional animation image segmentation.The test results show that this method can greatly remove the fog in the image,improve the clarity and color effect of the image.It can accurately complete the target image segmentation processing in the two-dimensional animation image,and obtain the complete target image.The consistency measure and inter-regional contrast of the segmented image are above 0.933,and the uniformity and contrast of the segmented image are good.The shape measurement results are all above 0.927,and the image quality after segmentation is good,which can meet the needs of two-dimensional animation image segmentation with different background complexity.