History – “Ancient” and Otherwise This week, our local historical museum took another step forward in making it relatively easy to explore local history. The Dept of Corrections in Oklahoma offered to scan our high school yearbook collection – for free – which allowed us to run what is known as “optical character recognition”, or OCR, on the resulting images. That’s technical talk for converting the images into actual text (letters, words, names, etc.), which then enables us to automatically search using words and names. Testing the results is important. Just how much of the text in the books did the system detect? How much of it did it correctly translate into the correct letters and words? How quickly can the collection be searched? How much hair did museum Board member Tim Quillen have when he . . . oh, wait, that fun part comes later! The results are not perfect, but very satisfying nonetheless. As yearbooks became more elegant, and printing techniques more advanced, there were a number of things done that put a challenge in the OCR work: 1) some pages have deeply colored backgrounds that fade from top to bottom – on those pages, the text is fairly easy for a human to decipher, but the machine misses some of the low contrast stuff near the top while it readily reads the higher contrast stuff near the bottom; 2) some type fonts are quite artsy-fartsy, but almost impossible for the text interpreter to decipher – words become jumbles of letters and special characters when the machine takes its best guess. But even with the omissions, the overall results are close to spectacular, when you consider that the alternative might be to read every book! Since I didn’t graduate from the local schools, I felt no fear in searching for Granite Falls friends. I typed in “Quillen”, and in 49 seconds our PC found 90 mentions of the name Quillen in 20 different yearbooks. It was tough to select a picture, but I chose the one from Tim’s senior year. But that immediately begged the question, “What about Tim’s son, Lakin, who graduated just a few years ago?” It was easy to simply click on Lakin’s name (toward the end of the list of 90) to find another curly-headed Quillen! Like his Dad, Lakin has done lots of great volunteer work at the museum, but given Tim’s current hairstyle, I hope Lakin doesn’t look at his Dad, read this, and make any projections  Tim ’79 Lakin ‘12 But another museum Board member is Ron Chew, whose farm was recently recognized for being in the same family for over 100 years. So I had to search on “Chew”, and was rewarded with 217 hits in 47 different yearbooks. Some of them were labeled “chewing gum”, “pencil chewing”, or in later years “Chewbacca”, but it took only seconds to spot Ron Chew in the Class of ’64. It was also fun to go back and look at pictures in the 1923 and 1924 yearbooks that included Wesley Chew! Ron ‘64 , OCR Text: History – “Ancient” and Otherwise This week, our local historical museum took another step forward in making it relatively easy to explore local history. The Dept of Corrections in Oklahoma offered to scan our high school yearbook collection – for free – which allowed us to run what is known as “optical character recognition”, or OCR, on the resulting images. That’s technical talk for converting the images into actual text (letters, words, names, etc.), which then enables us to automatically search using words and names. Testing the results is important. Just how much of the text in the books did the system detect? How much of it did it correctly translate into the correct letters and words? How quickly can the collection be searched? How much hair did museum Board member Tim Quillen have when he . . . oh, wait, that fun part comes later! The results are not perfect, but very satisfying nonetheless. As yearbooks became more elegant, and printing techniques more advanced, there were a number of things done that put a challenge in the OCR work: 1) some pages have deeply colored backgrounds that fade from top to bottom – on those pages, the text is fairly easy for a human to decipher, but the machine misses some of the low contrast stuff near the top while it readily reads the higher contrast stuff near the bottom; 2) some type fonts are quite artsy-fartsy, but almost impossible for the text interpreter to decipher – words become jumbles of letters and special characters when the machine takes its best guess. But even with the omissions, the overall results are close to spectacular, when you consider that the alternative might be to read every book! Since I didn’t graduate from the local schools, I felt no fear in searching for Granite Falls friends. I typed in “Quillen”, and in 49 seconds our PC found 90 mentions of the name Quillen in 20 different yearbooks. It was tough to select a picture, but I chose the one from Tim’s senior year. But that immediately begged the question, “What about Tim’s son, Lakin, who graduated just a few years ago?” It was easy to simply click on Lakin’s name (toward the end of the list of 90) to find another curly-headed Quillen! Like his Dad, Lakin has done lots of great volunteer work at the museum, but given Tim’s current hairstyle, I hope Lakin doesn’t look at his Dad, read this, and make any projections  Tim ’79 Lakin ‘12 But another museum Board member is Ron Chew, whose farm was recently recognized for being in the same family for over 100 years. So I had to search on “Chew”, and was rewarded with 217 hits in 47 different yearbooks. Some of them were labeled “chewing gum”, “pencil chewing”, or in later years “Chewbacca”, but it took only seconds to spot Ron Chew in the Class of ’64. It was also fun to go back and look at pictures in the 1923 and 1924 yearbooks that included Wesley Chew! Ron ‘64 , Granite Falls Historical Society,Documents (articles, clippings, letters, papers),Local History Articles,Other (articles, research, etc.),Ancient History.pdf,Ancient History.pdf Page 1, Ancient History.pdf Page 1