Line-Oriented+Handwriting+Analysis

1. The main reason of this page is to show us how forensic scientists compare in number of peoples handwriting, by looking at sentences or groups of words rather then one letter at a time. This also shows some scientific measurements on how to figure out the number of words per line or sentence.

2. The scientists compared a different number of peoples handwriting, looked at a different number of words, and looked at different letters of each word. The scientists came up with a recognition rate for each of the categories they compared.



3. We think that this experiment is good science because the chart shows a healthy trend in all of the data. For example, the experiment with 776 words has a recognition rate of 71.34% and as the number of words increase, the recognition rate decreases to show at 7719 words has a recognition rate of 60.05%.

4. The authors discovered that there is a comparision between handwritten sentence recognition. Methods used for single letter recognition were also used for sentence handwritting recognition. The author discovered that language is not a random character. They discovered this by handwritting recognition and line structure by experimenting and creating a chart to compare the number of words to the rate of recognition.

5. Interesting Facts: 1) It is interesting that they use a whole sentence instead of on work to do handwritting recongnition because it can change from time to time. 2) Database can be grouped into sets with different characteristics on the number of words in the vocabulary. 3) There are more verisons of this system where bigger and more vocabulary are used. 4) The first systems read segmented characters, and now there aiming at cursive writing. 5) This paper shows an experiment using many words instead of just looking at just one letter at a time and it has a better outcome. People don't write letters the same everytime so using more words is a bigger sample to look at and to have a better outcome.