findit

findit copied to clipboard

当模板图片比目标图片更大，IndexError

1

主要跟缩放机制有关

williamfzc

找到耗时有点长，能改进吗？

5

运行了你的demo,自己也试了自己的图片，发现：找到并返回结果 这个过程总耗时有点长（我本机电脑试了几次，都在3s+），能改进吗？

huage5

图片相似性

1

图片相似性已经部署了吗，想问怎么用

Neoyanghc

在保证准确率的前提下，这种方式能够提升分析效率

williamfzc

williamfzc

图片复杂度

1

针对黑白屏等异常情况的检测引擎

williamfzc

get text position with OCR engine

3

事实上 tesseract 提供了获取 box 的API，这使得获取文字所在位置成为可能。 tesserocr也提供了相应的API： ```python from PIL import Image import cv2 from tesserocr import PyTessBaseAPI, RIL IMAGE_PATH = r'tests/pics/screen.png' image = Image.open(IMAGE_PATH) cv2_image = cv2.imread(IMAGE_PATH) with PyTessBaseAPI(lang='eng+chi_sim') as...

williamfzc

Java: https://github.com/williamfzc/findit-java-client python: https://github.com/williamfzc/findit-client sync it with python version.

williamfzc

安装的时候报错了，希望能够支持aarch64，非常感谢大佬

2

[ 37%] Building C object Utilities/cmliblzma/CMakeFiles/cmliblzma.dir/liblzma/lzma/lzma_decoder.c.o [ 37%] Building C object Utilities/cmliblzma/CMakeFiles/cmliblzma.dir/liblzma/lzma/lzma_encoder.c.o [ 37%] Building C object Utilities/cmliblzma/CMakeFiles/cmliblzma.dir/liblzma/lzma/lzma_encoder_optimum_fast.c.o [ 37%] Building C object Utilities/cmliblzma/CMakeFiles/cmliblzma.dir/liblzma/lzma/lzma_encoder_optimum_normal.c.o [ 37%] Building C object Utilities/cmliblzma/CMakeFiles/cmliblzma.dir/liblzma/lzma/lzma_encoder_presets.c.o...

bcvb2001