Connected Component Analysis (CCA) is a powerful image processing technique used for the segmentation and extraction of useful information from images. It is especially useful for the study and analysis of objects present in images. In this assignment, we will use CCA to decode Braille language images. Braille is a tactile writing system used by visually impaired people, and its characters are formed using a combination of six raised dots arranged in a 3 × 2 matrix. Our goal is to develop a generic algorithm that can decode any image with Braille sequence to the corresponding English version by separating the Braille sequence from the background using 8-connectivity based CCA.
V={255} I am converting the gravyscale image to binary and then invert it. so the dots become bright and the background become dark.
The algorithm consist of the following steps:
To slice each row of the image, we can use the slicing operation in Python. This operation allows us to extract a specific portion of the image, in this case, a row, as a new image. Once we have sliced each row, we can apply the algorithm to each row individually to detect the dots.
To get the range of each dotted row, we can iterate over each row of the image and check for the presence of dots. We can define a threshold value for the intensity of the pixels that are considered as dots. If the intensity of a pixel is above this threshold, we consider it as a dot. Once we have detected the dots in each row, we can find the range of the dotted rows, i.e., the row numbers that contain dots.
To get the range of each dotted column, we can iterate over each column of the image and check for the presence of dots. Similar to the rows, we can define a threshold value for the intensity of the pixels that are considered as dots. If the intensity of a pixel is above this threshold, we consider it as a dot. Once we have detected the dots in each column, we can find the range of the dotted columns, i.e., the column numbers that contain dots.
To check each Braille position and make the Braille cell, we can use the range of dotted rows and columns that we obtained in steps 2 and 3. A Braille cell consists of six dots arranged in a 2x3 matrix. We can check each position of the Braille cell, i.e., whether it contains a dot or not, by comparing the row and column ranges with the position of each dot in the Braille cell.
To decode the Braille cell and obtain its corresponding alphabet, we can create a lookup table that maps each Braille cell to its corresponding alphabet. We can compare the Braille cell with this lookup table and obtain the corresponding alphabet. Once we have decoded all the Braille cells in the image, we can obtain the complete text.
