Electronic device and control method thereof

1. An electronic apparatus comprising:

a storage configured to store a training model of multi-dimensional long short-term memory (MDLSTM); and

a processor configured to acquire an image including at least one of handwritten text and printed text, identify each text line region of a plurality of text line regions in the image through image processing, recognize text included in the identified each text line region based on the training model, and identify a function related to the recognized text and perform the identified function,

wherein the processor is further configured to identify a text line region among the plurality of text line regions as a plurality of vertical blocks and calculate a probability of a character estimated from each of the plurality of vertical blocks based on a pixel value of each of the plurality of vertical blocks,

wherein the processor is further configured to combine an existence probability of a first character in a first vertical block of the plurality of vertical blocks, an absence probability of the first character in the first vertical block, an existence probability of a second character in a second vertical block of the plurality of vertical blocks, and an absence probability of the second character in the second vertical block to calculate a plurality of probabilities in which at least one of the first and second characters is included in the text line region,

wherein the processor is configured to binarize a text region in the image and perform the image processing by correction processing with respect to the binarized text region,

wherein the correction processing includes at least one of removal of a slope and skeletonization with respect to the text region,

wherein the processor is further configured to calculate an average value of a standardized text height through the image processing, apply a horizontal line of the calculated average value of the standardized text height and two horizontal lines spaced apart from the horizontal line to each of the plurality of text line regions in the image, and identify the each text line region of the plurality of text line regions based on the horizontal line and the two horizontal lines,

wherein one line of the two horizontal lines has a first interval from the horizontal line in a first vertical direction, and the other line of the two horizontal line has a second interval from the horizontal line in a second vertical direction, the first vertical direction being opposite to the second vertical direction, and the first interval being a same as the second interval,

wherein the plurality of text line regions in the image include a plurality of text lines, and

wherein the processor is further configured to individually identify each of the plurality of text lines, and identify text spanning the horizontal line and the two horizontal lines in the image as one text line of the plurality of text lines.

2. The apparatus of claim 1 , wherein the processor is configured to apply a token passing decoding and a preset language model to the plurality of calculated probabilities and acquire text recognized in the text line region.

3. The apparatus of claim 1 , further comprising:

a camera; and

a display,

wherein the image including at least one of the handwritten text and printed text refers to an image obtained by the camera or a handwritten image input on the display.

4. A method of controlling an electronic apparatus storing a training model of a multi-dimensional long short-term memory (MDLSTM), the method comprising:

acquiring an image including at least one of handwritten text and printed text;

identifying each text line region of a plurality of text line regions in the image through image processing;

recognizing text included in the identified each text line region based on the training model; and

identifying a function related to the recognized text and performing the identified function,

wherein the recognizing the text comprises identifying a text ling region of the plurality of text line regions as a plurality of vertical blocks and calculating a probability of a character estimated from each of the plurality of vertical blocks based on a pixel value of each of the plurality of vertical blocks,

wherein the calculating the probability comprises combining an existence probability of a first character in a first vertical block of the plurality of vertical blocks, an absence probability of the first character in the first vertical block, an existence probability of a second character in a second vertical block of the plurality of vertical blocks, and an absence probability of the second character in the second vertical block to calculate a plurality of probabilities in which at least one of the first and second characters is included in the text line region,

wherein the identifying each of the plurality of text line regions comprises binarizing a text region in the image and performing the image processing by correction processing with respect to the binarized text region,

wherein the correction processing includes at least one of removal of a slope and skeletonization with respect to the text region,

wherein the identifying the each text line region comprises calculating an average value of a standardized text height through the image processing, applying a horizontal line of the calculated average value of the standardized text height and two horizontal lines spaced apart from the horizontal line to each of the plurality of text line regions in the image, and identifying the each text line region of the plurality of text line regions based on the horizontal line and the two horizontal lines,

wherein one line of the two horizontal lines has a first interval from the horizontal line in a first vertical direction, and the other line of the two horizontal line has a second interval from the horizontal line in a second vertical direction, the first vertical direction being opposite to the second vertical direction, and the first interval being a same as the second interval,

wherein the plurality of text line regions in the image include a plurality of text lines,

wherein the identifying the each text line region comprises individually identifying each of the plurality of text lines by applying the horizontal line and the two horizontal lines to each of the plurality of text line regions in the image, and

wherein text spanning the horizontal line and the two horizontal lines in the image is identified as one text line of the plurality of text lines.

5. The method of claim 4 , wherein the recognizing the text comprises applying a token passing decoding and a preset language model to the plurality of calculated probabilities and acquiring text recognized in the text line region.

6. The apparatus of claim 1 , wherein the correction processing includes the skeletonization with respect to the text region, and

wherein the skeletonization is correction processing performed on the text region such that pixels forming text in the text region are densely formed by pixels of a first color, and pixels forming a region other than the text in the text region have a second color different from the first color.

7. The method of claim 4 , wherein the correction processing includes the skeletonization with respect to the text region, and

wherein the skeletonization is correction processing performed on the text region such that pixels forming text in the text region are densely formed by pixels of a first color, and pixels forming a region other than the text in the text region have a second color different from the first color.