Image processing using vector data to reduce noise

An image processing method comprising the steps of:

• inputting binary image data obtained by reading an image by means of an image reader;
• extracting contour-vector from the binary image data input in said inputting step;
• discriminating, based on the condition of the contour-vector data extracted in said extracting step, noise of the binary image data input in said inputting step, generated during the reading operation by the image reader;
• and
• rectifying the contour-vector data to delete the noise discriminated in said discriminating step,
• wherein, in a case where among five consecutive contour vectors, a contour vector situated at the center has length equal to 1, contour vector immediately before and after the contour vector at the center have lengths equal to 1 and directions opposite to each other, and second contour vectors before and after the contour vector at the center have lengths equal to at least 3 and the same direction as the contour vector at the center, the five consecutive contour vectors are rectified in said rectifying step.

An image processing apparatus comprising:

• input means for inputting image data;
• extraction means for extracting contour-vector data representing a contour defined between black pixels and white pixels of the image data, based on the image data input by said input means;
• and
• deletion means for deleting data corresponding to a closed loop made up of four vectors, each of which is one unit in length, and which are connected to each other, from among the contour-vector data extracted by said extraction means, wherein the data corresponding to the closed loop represents an isolated point in the image data.

An image processing apparatus comprising:

• input means for inputting image data;
• extraction means for extracting contour-vector data representing a contour defined between black pixels and white pixels of the image data, based on the image data input by said input means;
• and
• rectification means for rectifying data corresponding to, among the contour-vector data extracted by said extraction means, five consecutive vectors in which a vector situated at the center is one unit in length, vectors immediately before and after the vector at the center are each one unit in length, and second contour vectors before and after the contour vector at the center are each at least three units in length and extend in the same direction as the vector at the center.

An image processing apparatus comprising:

• input means for inputting image data;
• extraction means for extracting contour-vector data representing a contour defined between black pixels and white pixels of the image data, based on the image data input by said input means;
• and
• rectification means for rectifying data representing, among the contour-vector data extracted by said extraction means, seven consecutive vectors, each of which is one unit in length, and among which a target vector and second vectors before and after the target vector extend in the same direction, and the vectors immediately before and after the target vector are opposite in direction to each other, and the third vectors before and after the target vector are opposite in direction to each other.

An image processing method comprising the steps of:

• inputting image data;
• extracting contour-vector data representing a contour defined between black pixels and white pixels of the image data, based on the image data input in said inputting step;
• and
• deleting data corresponding to a closed loop made up of four vectors, each of which is one unit in length, and which are connected to each other, from among the contour-vector data extracted in said extracting step, wherein the data corresponding to the closed loop represents an isolated point in the image data.

An image processing method comprising the steps of:

• inputting image data;
• extracting contour-vector data representing a contour defined between black pixels and white pixels of the image data, based on the image data input in said inputting step;
• and
• rectifying data corresponding to, among the contour-vector data extracted in said extracting step, five consecutive vectors in which a vector situated at the center is one unit in length, vectors immediately before and after the vector at the center are each one unit in length, and second contour vectors before and after the contour vector at the center are each at least three units in length and extend in the same direction as the vector at the center.

An image processing method comprising the steps of:

• inputting image data;
• extracting contour-vector data representing a contour defined between black pixels and white pixels of the image data, based on the image data input in said inputting step;
• and
• rectifying data representing, among the contour-vector data extracted in said extracting step, seven consecutive vectors, each of which is one unit in length, and among which a target vector and second vectors before and after the target vector extend in the same direction, and the vectors immediately before and after the target vector are opposite in direction to each other, and the third vectors before and after the target vector are opposite in direction to each other.