Published Technical Disclosure Archive

AI-Powered Multi-Scale Disease Detection System

Publication ID: 24-11856881_0007_PTD
Published: October 26, 2025
Category:New Applications & Use Cases

Legal Citation

pr1or.art Inc., “AI-Powered Multi-Scale Disease Detection System,” Published Technical Disclosure No. 24-11856881_0007_PTD, Published October 26, 2025, available at https://archive.pr1or.art/24-11856881_0007_PTD
This technical disclosure describes improvements that would be readily apparent to a Person Having Ordinary Skill In The Art (PHOSITA) when considered in combination with the foundational architecture disclosed in U.S. Patent No. 11,856,881.

Background and Problem Solved

The original patent disclosed a system for detecting plant diseases using multi-stage, multi-scale deep learning. While this technology has proven effective in the agricultural domain, its potential applications extend far beyond. The present invention addresses the limitations of the original patent by exploring new industries and use cases where the core technology can provide significant value.

Novelty and Inventive Step

The new claims introduce novel applications and use cases for the core technology, which are not obvious from the original patent. The inventive step lies in the recognition of the technology's potential to address significant problems in diverse industries and the development of tailored systems and methods to achieve this.

Alternative Embodiments and Variations

Alternative embodiments of the invention could include adapting the core technology to other industries, such as healthcare, finance, or education. Variations of the invention could involve using different types of data, such as audio or text, or integrating the technology with other AI or machine learning approaches.

Potential Commercial Applications and Market

The present invention has significant commercial potential across various industries, including environmental monitoring, agriculture, insurance, logistics, and food safety. The target market includes companies, governments, and organizations seeking to leverage AI and machine learning to drive innovation and improve decision-making.

CPC Classifications

SectionClassGroup
A A01 A01B79/005
G G06 G06F18/214
G G06 G06F18/24317
G G06 G06F18/254
G G06 G06N3/045
G G06 G06N3/08
G G06 G06T3/40
G G06 G06T7/0012
G G06 G06V10/764
G G06 G06V10/774
G G06 G06V10/82
G G06 G06V20/188
G G06 G06V20/60
G G06 G06V20/68
A A01 A01G7/00
G G06 G06T2207/20016
G G06 G06T2207/20081
G G06 G06T2207/20084

Field of Art

Machine learning and computer vision for image classification and analysis, with specific focus on multi-stage deep learning models for detecting patterns in complex visual data across various domains

Person of Ordinary Skill (PHOSITA) Profile

A skilled practitioner with expertise in deep learning architectures, image processing, transfer learning techniques, and experience applying machine learning models to domain-specific classification problems

Obviousness Rationale

A PHOSITA would recognize that the core deep learning methodology for image classification disclosed in the source patent can be readily adapted to multiple domains by retraining the model on different image datasets. The multi-stage, multi-scale approach is fundamentally transferable across industries, with the primary innovation being domain-specific training data and targeted problem definition.

Obvious Combinations & Variations

Source Patent Element
Multi-stage deep learning classification model for detecting plant diseases
PTD Variation
Applying the same model architecture to detect water pollution through aquatic plant image analysis
Obviousness Reasoning
Predictable result of transferring a proven image classification technique to a related environmental monitoring problem, requiring only domain-specific training data
Source Patent Element
Image resizing and region extraction techniques for model input preprocessing
PTD Variation
Adapting preprocessing techniques to satellite imagery for crop yield prediction
Obviousness Reasoning
Known technique of applying consistent image preprocessing methods across different visual classification tasks, representing a standard machine learning approach
Source Patent Element
Deep learning model for detecting spatial patterns in complex visual data
PTD Variation
Extending the model to detect anomalies in property damage images for insurance fraud detection
Obviousness Reasoning
Obvious application of pattern recognition techniques to a different domain, leveraging the model's inherent capability to identify visual irregularities
Source Patent Element
Multi-scale classification approach for identifying discrete regions of interest
PTD Variation
Using similar techniques to detect defects in inventory management and supply chain images
Obviousness Reasoning
Finite set of known solutions for visual anomaly detection, representing a straightforward technological extension
35 U.S.C. § 103 Summary: Based on the teachings of US Patent 11856881, a person of ordinary skill in the art would find the disclosed variations obvious, as they represent predictable applications of a generalized deep learning image classification methodology to alternative domains, requiring only domain-specific training data and minor architectural adaptations.

Original Patent Information

Patent NumberUS 11,856,881
TitleDetection of plant diseases with multi-stage, multi-scale deep learning
Assignee(s)CLIMATE LLC