Text detection and classi!cation in natural images is important for many computer vision
applications. Here, to detect text, we !nd connected
components based on consistency in stroke width, chain
them based on their relative spatial positions, and use a
text classi!cation engine to !lter chains with low classi!cation con!dence scores. The algorithm performs
well using the ICDAR 2015 competition images, with
72.4% precision and 71.0% recall, but struggles under
certain imaging conditions or types of text. Finally, we
discuss ways to improve the algorithm to yield better
performance.
Text detection and classi!cation in natural images is important for many computer visionapplications. Here, to detect text, we !nd connectedcomponents based on consistency in stroke width, chainthem based on their relative spatial positions, and use atext classi!cation engine to !lter chains with low classi!cation con!dence scores. The algorithm performswell using the ICDAR 2015 competition images, with72.4% precision and 71.0% recall, but struggles undercertain imaging conditions or types of text. Finally, wediscuss ways to improve the algorithm to yield betterperformance.
การแปล กรุณารอสักครู่..