First, the scanning operation and points for attention Scanner online In the case that the "scanner test" can find the MCC scanner, you can run the OCR software of Shangshu No.6. Then click "Scan" button. A moment later, the scanner's control window appears, and the image is previewed with "black and white" at 300 dpi. The above steps can also be achieved by "OCR shortcut key" on the MCC scanner. At present, most MICOTEK scanners on the market are equipped with shortcut keys, which are convenient for users to use.
Enlarge the preview and adjust the sharpness of the image.
In order to achieve the best recognition effect, the minimum requirement for input manuscript during scanning is clarity. For this reason, we can sample and scan several characters in the manuscript through "Zoom in Preview" to adjust the brightness of the image more carefully. The adjustment tool is the "threshold" in the scanner tool.
The following are the scanning results under different thresholds. After adjusting to the appropriate threshold, you can select the Scan button. The scanning results will be transmitted to OCR software, and the control window of the scanner will disappear automatically.
Second, the matters needing attention before identification
After the above things are completed, what we need to do is the actual operation in OCR software.
Pay attention to the tilt correction of text.
Because the principle of OCR recognition is carried out in the form of fonts, we must pay attention to whether the manuscript is placed horizontally. In the specific implementation process, the image tilt correction button can be used to solve the problem.
Pretreatment of manuscript recognition
For the columns in the document, it is suggested that you set the recognition range manually, and it is best not to use "automatic segmentation" to ensure the consistency of recognition results.
Adopt appropriate identification methods.
In the specific recognition, you should also pay attention to whether your manuscript is horizontal or vertical, so as to choose the correct format button and keep the corresponding relationship.
At present, OCR software of Shangshu No.6 provides users with different recognition methods, such as simplified, traditional and English, and its choice is the drop-down menu on the window instead of the button menu. Simplified, Traditional and English buttons are the correct display modes of Shangshu No.6 on different operating systems, so don't confuse them.
After confirming the above steps, you can press the "Identify" button at this time. After recognition, the system will enter the "manuscript proofreading interface".
Third, manuscript proofreading
Generally speaking, OCR will display blue for the text that cannot be completely determined. Please confirm. However, it is worth noting that errors may occur where there are no errors, especially English words in Chinese texts. OCR generally recognizes them in Chinese, and the error rate is almost 100%. Therefore, when proofreading, we can read through it first to improve the proofreading effect.
We can add the text you need through the text input method provided by the operating system in this interface.
OCR provides the function of choosing an external editor, and we can choose a text editor.
Finally, it is suggested that you proofread and rearrange the whole manuscript in WORD to achieve the best effect of using OCR.