Keyword Detection Device Using Feature Point Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing key word detection technologies face challenges in rapidly and accurately extracting search key words from TV broadcasting images due to high processing burdens and long processing times, especially in environments requiring real-time detection and dealing with complex backgrounds or still images.
Innovation Solution
A key word detection device that rapidly detects search key words by identifying feature points of specific character strings in target images, restricting the search window to a rectangular shape, and recognizing character strings within these windows, thereby reducing processing time and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If character recognition processing is performed on the entire TV broadcasting image, then the search key word can be extracted, but the processing time and processing burden increase significantly
Solution Approach 1:
The patent divides the character recognition task into two segments: first detecting feature points of specific character strings (such as search prompts), then performing character recognition only in the surrounding regions of these feature points. This segmentation approach reduces the processing area from the entire image to only relevant regions, thereby reducing processing time while maintaining detection accuracy.
Solution Approach 2:
The patent performs preliminary feature point detection on specific character strings before performing character recognition. By pre-identifying the locations of search prompts and other specific characters, the system can then focus character recognition resources only on these pre-identified regions, avoiding unnecessary processing of the entire image and thus reducing overall processing time.
2Adaptability or versatility
If character recognition is performed on the entire image, then all possible key words can be detected, but the processing burden and memory requirements increase
Solution Approach 1:
The patent applies local quality by focusing character recognition processing only on specific local regions surrounding feature points of interest (such as search prompts). Instead of uniformly processing the entire image, the system concentrates computational resources on localized areas where key words are likely to appear, thereby reducing processing burden while maintaining comprehensive detection capability.
Solution Approach 2:
The patent extracts and processes only the relevant portions of the image by first identifying feature points of specific character strings and then performing character recognition only in the surrounding regions. This extraction approach removes unnecessary processing from irrelevant areas of the image, reducing overall processing burden while preserving the ability to detect all possible key words.
3Productivity
If the search key word is displayed for a short time in TV broadcasting, then the advertising effectiveness is improved, but the user may miss the search key word
Solution Approach 1:
The patent enables the system to automatically detect and extract search key words from TV broadcasting images without requiring user intervention. By using feature point detection and localized character recognition, the system serves itself by autonomously identifying and providing search queries, ensuring reliable key word recognition even when displayed for short periods, thereby maintaining both advertising effectiveness and recognition reliability.
Data Source
AI summary
A key word detection device and a method for detecting a search key word from a target image in order to perform a search with a search engine on the internet, the key word detection device comprising: a processor configured to operate as a feature point detector configured to detect a feature point of a specific character string from the target image, the specific character string prompting a user to perform the search; a key word recognition unit configured to recognize a character string existing in surroundings of the feature point detected by the feature point detector as the search key word in the target image; and a storage for storing character information and data of the target image used by the processor.