颜色差信息提取耦合圆拟合技术的果树桃子检测算法

Peach Detection Algorithm Based on Information Extraction of Color Difference Coupled Circle Fitting Technique

机器人在采摘果树桃子果实时,需要从复杂的果树背景中识别出桃子,以此为依据计算出桃子的圆心坐标与半径,传送给机器人系统进行作业,其中果树桃子的准确检测是机器人采摘作业的关键步骤。对此,提出了颜色差信息提取耦合圆拟合技术的果树桃子检测算法。首先利用颜色差信息提取桃子区域;然后基于边界跟踪处理与匹配膨胀处理得到桃子的完整区域;再分别用圆拟合方法计算出桃子的圆心,完成定位检测。该算法的性能测试结果表明:与当前水果检测算法相比,该算法具有更好的检测效果,可准确定位出桃子的目标位置。

When robot is picking peach, robot needs to identify it from the complicated background of fruit trees, and on this basis to calculate the peach center coordinates and radius and transmit to the robot system for operation. Due to the work environment that tends to have interference such as strong light, insufficient light or shade leaves, and the fruit trees peach accurate detection is the key step in the robot＇ s picking operation. To this, this paper proposed a fruit peach detection mechanism based on extraction information of color difference coupled circle fitting technique. First using color differ- ence information extraction peach area, and then treatment with matching expansion based on bounda- ry tracing, complete area of peach was got, and again with this circle fitting method, the center of the circle of peach to fimsh the fruit detection was calculated. The results of the tested performance of mechanism show that compared with other fruit detection mechanism, this mechanism has better detec- tion effect and can accurately pinpoint the peach target location.