试验采用均匀设计，在大田条件下固定数码相机高度垂直拍摄夏玉米拔节期和大喇叭口期的群体图像，利用图像处理技术获取玉米地面覆盖度，建立覆盖度与人工测得的叶面积指数(LAI)和干物质积累(DMA)的回归关系模型，并对该模型的适用性进行统计检验。结果表明：地面覆盖度与LAI和DMA间存在极显著正相关关系，相关系数分别达到了0.946和0.935，在叶面积估算模型中引入密度因素后模型的精确性得到了进一步改善，表明利用图像处理技术估测夏玉米群体长势具有很好的可行性。

The test adopted uniform design method,under field conditions,the group images of summer maize were obtained vertically by a digital camera from fixed height at jointing stage and bell stage.Then the test established the relation models between GCR extracted via image processing technique and leaf area index(LAI),dry matter accumulation(DMA),which were measured by traditional methods,and the practicality of those models were tested.The results showed that LAI and DMA were significantly correlated with GCR,and the determination coefficient(R²) were 0.946 and 0.935,respectively.And further analysis found that multiple regression model by adding density factor can effectively improve precision of LAI prediction model.Generally,the test results indicated that it has better feasibility to monitoring the growth of summer maize via using image processing technique.

国家“863”课题“精准农业生产设计与管理决策模型技术研究”(2006AA10A303)