Features:

  •  RecognitionP_PHA、RecognitionT_PHA —-Perceptual hash algorithm function, the parameters are the image disk path and Txture2D.
  •  RecognitionP_PSNR、RecognitionT_PSNR—-Peak Signal to Noise Ratio function, the parameters are the image disk path and Txture2D.
  •  RecognitionP_MSSIM、RecognitionT_MSSIM—-Structural similarity function, the parameters are the image disk path and Txture2D.
  • LoadFileToTexture2D—-Read the file from the disk to Texture2D.
  • SaveTexture2DToFile—-Save Texture2D to disk file.

Code Modules:

  •  ImageSimilarityRecognition [Runtime]

Check out ywt20's other products here : https://www.unrealengine.com/marketplace/en-US/profile/YWT20

Number of Blueprints: 1

Number of C++ Classes: 2

Network Replicated: (No)

Supported Development Platforms: (“Windows”)

Supported Target Build Platforms: ("Win64", "Win32")

Documentation: Link

Example Project: Link

Important/Additional Notes:Please note that if you want to use the Texture2D file included in the project, we need to set the value of CompressionSettings to UserInterface2D and the value of MipGenSettings to NoMipmaps. If it is a picture file loaded from a disk, there is no problem.

特征:

  •  RecognitionP_PHA、RecognitionT_PHA—-感知哈希算法函数，参数为镜像盘路径和Txture2D。
  •  RecognitionP_PSNR、RecognitionT_PSNR—-峰值信噪比函数，参数为映像盘路径和Txture2D。
  •  RecognitionP_MSSIM、RecognitionT_MSSIM—-结构相似性函数，参数为镜像盘路径和Txture2D。
  • LoadFileToTexture2D—-从磁盘读取文件到Texture2D。
  • SaveTexture2DToFile—-保存Texture2D到磁盘文件。

代码模块:

  •  ImageSimilarityRecognition[运行时]

在这里查看ywt20的其他产品 : https://www.unrealengine.com/marketplace/en-US/profile/YWT20

蓝图数目:1

C++类数：2

网络复制：（否）

支持的开发平台：（”Windows”）

支持的目标构建平台：（”Win64″，”Win32″）

文件: 连结

示例项目: 连结

重要/附加注意事项:请注意，如果要使用项目中包含的Texture2D文件，我们需要将CompressionSettings的值设置为UserInterface2D，将MipGenSettings的值设置为NoMipmaps。 如果是从磁盘加载的图片文件，则没有问题。










My other plug-in “Similarity Recognition” is mainly for objects such as shapes and symbols.

In UE5, you need to add plugin content in Additional Asset Directories to Cook in the project settings. Otherwise, it will crash when used after packaging. And when loading pictures, try to use PNG pictures, JPG still has problems that have not been solved.

Open the Content Browser window, in the upper right corner of the Content Browser, click the Settings button, This will open a menu where you can adjust various settings for the current instance of the Content Browser, Under the View category, check the Show Plugin Content option, Find the Image Similarity Recognition->Content->Image Similarity Recognition directory, There are sample levels and related assets inside.

Several recognition functions are implemented in this plug-in based on OpenCV:

 PHA (perceptual hash algorithm). Its function is to generate a “fingerprint” string for each image, and then compare the fingerprints of different images. The closer the results, the more similar the images.

 PSNR (Peak Signal to Noise Ratio). It is a full reference image quality evaluation index. The most common and widely used image objective evaluation index, but it is based on the error between corresponding pixels, that is, based on error-sensitive image quality evaluation. Since the visual characteristics of the human eye are not considered (the human eye is more sensitive to the contrast difference with lower spatial frequency, the human eye is more sensitive to the brightness contrast difference than the chromaticity, and the human eye’s perception of an area will be affected. The impact of its surrounding neighboring areas, etc.), so the evaluation results often appear inconsistent with people’s subjective feelings.

 SSIM (structural similarity). It is a structural similarity and a full-reference image quality evaluation index. It measures image similarity from three aspects: brightness, contrast, and structure.

Change log : (2022/04)

Support 5.0 version

我的另一个插件”相似性识别“主要用于形状和符号等对象。

在UE5中，您需要在其他资源目录中添加插件内容，以便在项目设置中进行烹饪。 否则，包装后使用时会崩溃。 而在加载图片的时候，尽量使用PNG图片，JPG还是有问题没有解决。

打开内容浏览器窗口，在内容浏览器的右上角，单击设置按钮，这将打开一个菜单，您可以在其中调整内容浏览器当前实例的各种设置，在视图类别下，检查显示插件内容选项，找到图像相似性识别->内容->图像相似性识别目录，里面有样本级别和相关资产。

在这个基于OpenCV的插件中实现了几个识别功能:

 PHA（感知哈希算法）。 它的功能是为每个图像生成一个”指纹”字符串，然后比较不同图像的指纹。 结果越接近，图像越相似。

 PSNR(峰值信噪比）。 为全参考图像质量评价指标。 最常见、应用最广泛的图像客观评价指标，但它是基于对应像素之间的误差，即基于误差敏感的图像质量评价。 由于没有考虑人眼的视觉特性（人眼对空间频率较低的对比度差比较敏感，人眼对亮度对比度差比色度更敏感，人眼对一个区域的感知会受到影响。 其周边邻近地区的影响等。），因此评价结果往往出现与人的主观感受不一致的情况。

 SSIM（结构相似性）。 为结构相似性和全参考图像质量评价指标。 它从亮度、对比度和结构三个方面来衡量图像相似性。

更改日志 : (2022/04)

支持5.0版本

声明：本站所有资源都是由站长从网络上收集而来，如若本站内容侵犯了原著者的合法权益，可联系站长删除。