In multispectral transmission imaging of biological tissues, it is crucial to increase the grayscale resolution for accurate heterogeneity detection. This study proposes a method of obtaining high-grayscale-resolution images via image oversampling and downsampling. A single-wavelength transmission image of a simulated heterogeneous breast tissue was used to validate the effectiveness of the method experimentally. In the experiment, a high-resolution camera was used to capture transmission images of the simulated tissue with high spatial resolution $${\text{I}}_{\lambda }^{h}$$ I λ h and low spatial resolution $${\text{I}}_{\lambda }^{l}$$ I λ l . The high-spatial-resolution image $${\text{I}}_{\lambda }^{h}$$ I λ h was compressed into an image $${\text{I}}_{\lambda }^{hc}$$ I λ hc with the same spatial resolution as that of the low-spatial-resolution image $${\text{I}}_{\lambda }^{l}$$ I λ l , thereby converting the spatial information into image depth information. The experimental results indicate that the compressed image $${\text{I}}_{\lambda }^{hc}$$ I λ hc had higher grayscale resolution and richer grayscale details than those of the low-spatial-resolution image $${\text{I}}_{\lambda }^{l}$$ I λ l . Our results prove that it is feasible to improve the grayscale resolution of an image through image oversampling and downsampling while sacrificing the image spatial resolution to a certain extent. Further, this research provides an effective means to analyze heterogeneity in multispectral transmission imaging of a biological tissue.

中文翻译：

---

采用图像过采样和下采样技术来提高灰度分辨率

在生物组织的多光谱透射成像中，提高灰度分辨率对于准确的异质性检测至关重要。本研究提出了一种通过图像过采样和下采样获得高灰度分辨率图像的方法。模拟异质乳腺组织的单波长透射图像用于通过实验验证该方法的有效性。在实验中，使用高分辨率相机拍摄模拟组织的透射图像，具有高空间分辨率$${\text{I}}_{\lambda }^{h}$$ I λ h 和低空间分辨率$${\text{I}}_{\lambda }^{l}$$ I λ l 。高空间分辨率图像 $${\text{I}}_{\lambda }^{h}$$ I λ h 被压缩成图像 $${\text{I}}_{\lambda }^ {hc}$$ I λ hc 具有与低空间分辨率图像 $${\text{I}}_{\lambda }^{l}$$ I λ l 相同的空间分辨率，从而转换空间信息转化为图像深度信息。实验结果表明，压缩后的图像$${\text{I}}_{\lambda }^{hc}$$ I λ hc 比低空间分辨率的图像具有更高的灰度分辨率和更丰富的灰度细节。 ${\text{I}}_{\lambda }^{l}$$ I λ l 。我们的结果证明，在一定程度上牺牲图像空间分辨率的同时，通过图像过采样和下采样来提高图像的灰度分辨率是可行的。更多，