Change detection in boreal forests using bi-temporal aerial photographs

Abstract

Increased need for timely forest information is leading to continuous updating of stand databases. In continuous updating, stand attributes are estimated in the field after an operation and stored in databases. To find the changes caused by operations and forest damage, a semi-automatic method based on bi-temporal aerial photographs was developed. The test data were classified into three classes: No-change (952 stands), Moderate-change (163 stands) and Considerable-change (44 stands). The aerial photographs were acquired in years 2001 and 2004 with almost the same image specifications. Altogether 110 features at stand level were extracted and used in change detection analysis. The test data were classified with stepwise discriminant analysis. The overall accuracy of classification varied between 75.3 and 84.7%. The considerable changes were found without error, whereas the Moderate-change and No-change classes were often confused. However, 84.2% of thinned stands were classified correctly. The best accuracy in classification was obtained by using the histogram and textural features extracted from the original, uncorrected images. Radiometric correction did not improve the accuracy of classification. Soil type, characteristics of the growing stock and the location of a stand in an image were found to affect the change detection. Before the method can be applied operationally, issues related to, e.g., confusion between No-change and Moderate-change must be solved.

Change detection using bi-temporal aerial photographs and registration at the stand level

“…Radiometric correction is usually required in order to decrease the effect of the bi-directional effect. However, in the study of Hyvönen and Anttila (2006), the radiometric correction of aerial photographs did not improve the classification accuracy. In that study, the aerial photographs were taken with the same image specifications at two time points, so the unchanged stands look similar in both aerial photographs.…”

“…These aerial photographs were used as base photographs for the extraction of spectral features. In a study by Hyvönen and Anttila (2006), it was discovered that radiometric correction did not significantly improve the classification results in the case of similar image specifications. In our case, the time gap between aerial photographs was less than half an hour.…”

“…Also, the study of Saksa et al (2003) revealed that distinct changes, such as clear cuttings, can be found with methods suitable for operational use. In the study of Hyvönen and Anttila (2006), clear changes were detected without error, and some improvement in detection of moderate changes was achieved, e.g. 84% of thinned stands were detected.…”

“…in the timber market and in timber prices, up-to-date databases are essential. Thus, there is an increased need for timely forest information also in private forestry, which is leading to a continuous updating of private forest owners' stand databases (Hyvönen and Anttila 2006).…”

Estimating storm damage with the help of low-altitude photographs and different sampling designs and estimators

“…If only photographs taken after the storm are used, it might be difficult to interpret whether the damage is due to a storm or not. Bi-temporal (or multi-temporal) photographs and other remote sensing materials enable the use of automatic image interpretation of storm damage and even the estimation of damage level (Hyvönen and Anttila 2006;Vastaranta et al 2012;Honkavaara et al 2013;Einzmann et al 2017). Using laser scanning data, a digital surface model (DSM) can be produced after the storm and holes in the canopy can be identified (Shedd et al 2006) or DSM's based on data scanned both before and after the storm can be compared (Vastaranta et al 2012;Honkavaara et al 2013).…”