A deep learning model-based plankton image enhancement method

By employing a three-level scale resolution and pixel-by-pixel phase alignment calibration through a deep learning model, combined with texture and color feature optimization, the shortcomings of texture resolution and color restoration in plankton image enhancement are addressed, achieving high-quality image processing and meeting subsequent application requirements.

CN121937320BActive Publication Date: 2026-07-21FIRST INSTITUTE OF OCEANOGRAPHY MNR
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
FIRST INSTITUTE OF OCEANOGRAPHY MNR
Filing Date
2025-12-18
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing plankton image enhancement techniques struggle to achieve accurate multi-scale texture analysis and lack effective phase calibration and feature optimization, resulting in low recognition and processing efficiency in densely textured areas. In terms of color restoration, they fail to establish a precise correspondence between pigment spectral absorption characteristics and color channels, leading to monotonous color levels in pigment areas. In overall processing, the fusion of texture and color areas lacks precise coordinate matching, resulting in abrupt transitions between enhanced and original areas, which damages image integrity and harmony.

Method used

A deep learning-based approach is employed to extract the spatial arrangement features of plankton texture and the spectral absorption features of chlorophyll-dominated pigment regions through three-level scale analysis. Pixel-by-pixel phase alignment calibration and feature overlay optimization are then performed. By precisely matching the coordinate matrices of the texture optimization region and the color calibration region, the consistency of feature fusion is ensured. Furthermore, the color ratio is adjusted by dynamically allocating channel weights to create a natural transition.

Benefits of technology

It achieves clear and regular texture features and true color reproduction in plankton images, improves texture recognition and color differentiation, ensures the overall coordination and integrity of the image, and meets the high-quality application requirements of subsequent morphological recognition and species classification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121937320B_ABST
    Figure CN121937320B_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of image processing, and particularly relates to a plankton image enhancement method based on a deep learning model. The method comprises the following steps: obtaining an original image of plankton to be enhanced; analyzing the original image through a deep learning model to extract a periodic texture spatial arrangement feature of a plankton shell pattern, a flagellum, and a spectral absorption feature of a pigment region dominated by chlorophyll; positioning a texture dense region based on the texture spatial arrangement feature, performing pixel-by-pixel phase alignment calibration on the texture dense region, and optimizing the texture feature of the texture dense region through feature superposition to form a texture optimization region. The present application realizes accurate optimization of plankton image texture features and natural restoration of color properties through a deep learning model and image processing enhancement technology, so as to improve the detail recognition, color authenticity and overall integrity of the plankton image.
Need to check novelty before this filing date? Find Prior Art