Multi-crop-row detection based on strip analysis

A method based on strip analysis was put forward to detect multiple crop rows from farmland features images. The image is divided into horizontal image strips. After points indicating center of the rows of each strip was detected, these points are classified by rows. The estimation of the position of center line of the rows is accomplished by least squares regression analysis. Experiments have proved the proposed method to be a reliable crop row detection method which can detect multiple crop rows with the advantage of small size of computer memory and short computational time. The accuracy of the estimation was determined by comparing the calculated row center line with the manual detected row position.