Use of error matrices to improve area estimates with maximum likelihood classification procedures

Abstract The maximum likelihood classifier is by far the most widespread among supervised classification methods. This procedure offers numerous advantages, but it has considerable shortcomings in the presence of strongly irregular spectral distributions, mainly related to bias in area estimates. Since these cases are quite common, some methods have already been proposed to correct biased area estimates from maximum likelihood classifications, but they are often not generally applicable or statistically stable. In this article a method is put forward to correct maximum likelihood assignment probabilities by means of a transition matrix; this matrix is derived through a simple mathematical algorithm from a contingency table of a previous classification compared to reference pixels. The purpose is clearly to attain a diagonalization of the final error sources to better estimate area extents and, above all, to achieve higher global discrimination accuracy. As different environmental situations may cause wide variability in the results of such a procedure, this was tested in three case studies using TM data acquired over areas with different landscapes. The results, evaluated by means of suitable statistics, significantly support that the method has general validity and applicability.

[1]  G. H. Rosenfield,et al.  A coefficient of agreement as a measure of thematic classification accuracy. , 1986 .

[2]  J. L. Smith,et al.  Using classification error matrices to improve the accuracy of weighted land-cover models , 1987 .

[3]  Paul J. Curran,et al.  Simple size for ground and remotely sensed data , 1986 .

[4]  Fabio Maselli,et al.  Use of multitemporal information to improve classification performance of TM scenes in complex terrain , 1991 .

[5]  Alan H. Strahler,et al.  The Use of Prior Probabilities in Maximum Likelihood Classification , 1980 .

[6]  Russell G. Congalton,et al.  A review of assessing the accuracy of classifications of remotely sensed data , 1991 .

[7]  Russell Congalton,et al.  A Review of Three Discrete Multivariate Analysis Techniques Used in Assessing the Accuracy of Remotely Sensed Data from Error Matrices , 1986, IEEE Transactions on Geoscience and Remote Sensing.

[8]  F. J. Ahern,et al.  Forestry information content of Thematic Mapper data , 1986 .

[9]  A. M. Hay,et al.  The derivation of global estimates from a confusion matrix , 1988 .

[10]  D. J. Booth,et al.  A comparison of classification algorithms in terms of speed and accuracy after the application of a post-classification modal filter , 1989 .

[11]  F. Maselli,et al.  Forest classification by principal component analyses of TM data , 1988 .

[12]  R. G. Oderwald,et al.  Assessing Landsat classification accuracy using discrete multivariate analysis statistical techniques. , 1983 .

[13]  Jacob Cohen.  A Coefficient of Agreement for Nominal Scales , 1960 .

[14]  F. Maselli,et al.  Inclusion of prior probabilities derived from a nonparametric process, into the maximum-likelihood classifer , 1992 .

[15]  K. Hall-Könyves.  Crop monitoring in Sweden , 1990 .

[16]  Fabio Maselli,et al.  Use of error probabilities to improve area estimates based on maximum likelihood classifications , 1990 .

[17]  Alan M. Hay.  Global estimates from confusion matrices A reply to Jupp , 1989 .

[18]  P. Curran.  Principles of remote sensing , 1984 .

[19]  David L.B. Jupp,et al.  The stability of global estimates from confusion matrices , 1989 .

[20]  Frank W. Davis,et al.  Performance analysis of image processing algorithms for classification of natural vegetation in the mountains of Southern California , 1986 .

[21]  T. M. Lillesand,et al.  Rapid maximum likelihood classification , 1991 .

[22]  F. Sabins.  Remote Sensing: Principles and Interpretation , 1987 .

[23]  A. Skidmore.  An expert system classifies eucalypt forest types using thematic mapper data and a digital terrain model , 1989 .

[24]  N. Fuhs,et al.  Evaluation of several schemes for classification of remotely sensed data , 1980 .

[25]  R. Congalton,et al.  Accuracy assessment: a user's perspective , 1986 .