C语言基础：理解LongCat-Image-Edit V2底层图像处理算法

代码import cv2 import numpy as np import time def super_sample(file, scale, count): print print(file, scale, count) # 读取图像 img cv

imread(file) start_time time.time() for index in range(count): # 双三次插值 bicubic cv

resize(img, None, fxscale, fyscale, interpolationcv

INTER_CUBIC) print(CUBIC cost time: %.2f ms % ((time.time()-start_time)*1000/count)) # cv

imwrite(cv2-CUBIC.jpg, bicubic) start_time time.time() for index in range(count): # Lanczos插值 lanczos cv

resize(img, None, fxscale, fyscale, interpolationcv

INTER_LANCZOS

print(LANCZOS4 cost time: %.2f ms % ((time.time()-start_time)*1000/count)) # cv

imwrite(cv2-LANCZOS

jpg, lanczos) count 1000 file1 1280x

png super_sample(file1, 2, count) super_sample(file1, 3, count) file2 1920x

png super_sample(file2, 2, count) super_sample(file2, 3, count)测试结果在虚拟机中测试的。

1280x7202x23x3CUBIC

2.

1

83LANCZOS

410.

4

891920x10802x23x3CUBIC

4.

4

00LANCZOS

424.

2

56

17.c-起草网登录免费-17.c-起草网登录免费应用