![ZEISS Sigma 300を用いた軟磁性複合材料の量的EBSD観察 ZEISS Sigma 300を用いた軟磁性複合材料の量的EBSD観察]({"xsmall":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.100.100.230,0,824,594.file/screensaver-routine-metallography-3.jpg","small":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.360.360.230,0,824,594.file/screensaver-routine-metallography-3.jpg","medium":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.594.594.230,0,824,594.file/screensaver-routine-metallography-3.jpg","large":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.594.594.230,0,824,594.file/screensaver-routine-metallography-3.jpg","xlarge":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.594.594.230,0,824,594.file/screensaver-routine-metallography-3.jpg","xxlarge":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.594.594.230,0,824,594.file/screensaver-routine-metallography-3.jpg","max":"https://www.zeiss.com/content/dam/rms/reference-master/resource-center/insights-hub/thumbnails/screensaver-routine-metallography-3.jpg/_jcr_content/renditions/original.image_file.594.594.230,0,824,594.file/screensaver-routine-metallography-3.jpg"})
金属組織検査のルーチンワーク用顕微鏡ソリューション
金属組織検査のルーチンに革命をもたらす
最先端の金属分析用顕微鏡ソリューション
金属の特性と使用中のその挙動は、金属の微細構造に左右されます。また、重要な機械的特性、腐食挙動、および疲労性能のすべては、粒径、組成、相の大きさ・分布、介在物、および局在する微細構造の変異の組み合せに影響を受けます。
ZEISSは、幅広い顕微鏡法に対応する、金属組織検査および品質管理のための最先端ソリューションのポートフォリオを提供しています。
![機械学習を用いて画像解析を改善 機械学習を用いて画像解析を改善]({"xsmall":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.100.67.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","small":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.360.240.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","medium":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.768.512.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","large":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.960.640.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","xlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.960.640.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","xxlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.960.640.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","max":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.960.640.0,65,960,705.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg"})
![機械学習を用いて画像解析を改善 機械学習を用いて画像解析を改善]({"xsmall":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.100.80.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","small":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.360.289.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","medium":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original.image_file.768.616.file/rapid-machine-learning-segmentation-of-duplex-ss.jpg","large":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original./rapid-machine-learning-segmentation-of-duplex-ss.jpg","xlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original./rapid-machine-learning-segmentation-of-duplex-ss.jpg","xxlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original./rapid-machine-learning-segmentation-of-duplex-ss.jpg","max":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/rapid-machine-learning-segmentation-of-duplex-ss.jpg/_jcr_content/renditions/original./rapid-machine-learning-segmentation-of-duplex-ss.jpg"})
機械学習を用いて画像解析を改善
画像や3Dデータセットの定量的解析には、必ずセグメンテーションステップが必要です。セグメンテーションとは、画像を複数の領域に分割し、何らかの方法でそれぞれを識別する操作です。セグメンテーション完了後、各領域を解析してアクションにつながる実際のデータを取得します。領域には、個別の粒子や介在物、孔、また様々な相や層等の確認が必要な対象物が含まれます。
セグメンテーションが難しい領域の例
- 色やコントラストが似ている領域
- テクスチャだけが異なる領域
- 紛らわしいアーティファクトやスクラッチ
- 3Dデータセットのノイズにより、セグメンテーションの正確性が損なわれる場合
- マルチチャンネル(RGBまたはそれ以上)画像の解析が複雑な場合
![ZEISS ZEN Intellesisは、ガイド付き機械学習を用いてこのようなセグメンテーションの問題に対処しています。 ZEISS ZEN Intellesisは、ガイド付き機械学習を用いてこのようなセグメンテーションの問題に対処しています。]({"xsmall":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.100.75.0,51,960,771.file/zen-intellesis-monitor_2021.jpg","small":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.360.270.0,51,960,771.file/zen-intellesis-monitor_2021.jpg","medium":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.768.576.0,51,960,771.file/zen-intellesis-monitor_2021.jpg","large":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.960.720.0,51,960,771.file/zen-intellesis-monitor_2021.jpg","xlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.960.720.0,51,960,771.file/zen-intellesis-monitor_2021.jpg","xxlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.960.720.0,51,960,771.file/zen-intellesis-monitor_2021.jpg","max":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.960.720.0,51,960,771.file/zen-intellesis-monitor_2021.jpg"})
![ZEISS ZEN Intellesisは、ガイド付き機械学習を用いてこのようなセグメンテーションの問題に対処しています。 ZEISS ZEN Intellesisは、ガイド付き機械学習を用いてこのようなセグメンテーションの問題に対処しています。]({"xsmall":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.100.104.file/zen-intellesis-monitor_2021.jpg","small":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.360.375.file/zen-intellesis-monitor_2021.jpg","medium":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original.image_file.768.800.file/zen-intellesis-monitor_2021.jpg","large":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original./zen-intellesis-monitor_2021.jpg","xlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original./zen-intellesis-monitor_2021.jpg","xxlarge":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original./zen-intellesis-monitor_2021.jpg","max":"https://www.zeiss.com/content/dam/rms/reference-master/applications/raw-materials-industrial-r-d/metals/zen-intellesis-monitor_2021.jpg/_jcr_content/renditions/original./zen-intellesis-monitor_2021.jpg"})
ZEISS ZEN Intellesisは、ガイド付き機械学習を用いてこのようなセグメンテーションの問題に対処しています。シンプルなグラフィックワークフローを採用したZEISS ZEN Intellesisは、ユーザーから提供されたトレーニング用入力データを使用して、30を超える各種パラメータを用いて一つひとつピクセルを評価し、正しいカテゴリーに割り当てます。このプロセスを反復することで、機械学習モデルをトレーニングし、検出能力と正確性を向上させます。その後、機械学習がデータセット全体に自動で適用されて、数百の画像や3Dデータセットが分析可能なフォーマットにセグメンテーションされます。機械学習についての知識や経験がないユーザーでも、ZEISS ZEN Intellesisを使えば日々のタスクにAIのパワーを活用することができます。