Image Processing Device Shape Color Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing methods for identifying unknown plants are inefficient in narrowing down similar images from a database, particularly when the target image has diverse colors or shapes, leading to difficulties in accurate identification.
Innovation Solution
An image processing device and method that acquires an image, compares its shape and color information with a database, selects representative images based on similarity, and displays them alongside the target image to facilitate identification, using a primary selection by shape and secondary selection by color.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image comparison methods are used to identify unknown plants, then the system can process images from a database, but the efficiency is low and accuracy is poor when dealing with diverse colors and shapes
Solution Approach 1:
The patent segments the image comparison process into two distinct stages: first comparing shape information to identify candidate types, then comparing color information to select the final match. This segmentation allows each comparison stage to focus on specific features, improving both accuracy and efficiency by avoiding the need to compare all images against all features simultaneously.
Solution Approach 2:
The patent changes the parameters used for comparison at different stages: shape parameters (such as contour, area, perimeter) are used in the first comparison stage, while color parameters (such as hue, saturation, brightness) are used in the second stage. This parameter change strategy enables the system to handle diverse plant variations effectively by adapting the comparison criteria to the specific characteristics being evaluated.
2Reliability
If all candidate images are displayed for comparison, then the user can see all possibilities, but the user interface becomes complex and difficult to operate
Solution Approach 1:
The patent segments the result display into two organized groups: candidate plant types from shape comparison and final match from color comparison. This segmentation presents information in a hierarchical manner that is easy to understand, showing users the progression from potential matches to the most likely identification without overwhelming them with all possible images simultaneously.
Solution Approach 2:
The patent extracts and displays only the most relevant information for user decision-making: the candidate types based on shape and the single best match based on color. By taking out only the essential comparison results rather than displaying all database images, the system maintains high identification reliability while keeping the interface simple and focused on what matters most to the user.
Data Source
AI summary
The image acquisition unit 41 acquires an image including an object. By comparing information related to the shape of a relevant natural object that is included as the object in the target image acquired by the image acquisition unit 41, and information related to respective shapes of a plurality of types prepared in advance, at least one flower type for the natural object in question is selected. The secondary selection unit 43 then selects data of a representative image from among data of a plurality of images of different color, of the same flower type as prepared in advance, for each of at least one flower type selected by the primary selection unit 42, based on information related to color of the relevant natural object included as the object in the image acquired by the image acquisition unit 41.