feat: SIMD OCR preprocessing — binarization, adaptive threshold, deskew

ndarray/hpc/ocr_simd.rs: SIMD-accelerated image preprocessing for OCR:
  - Otsu binarization: U8x64 histogram + optimal threshold
  - Bit-packed BinaryImage: 64 pixels per u64 word
  - Adaptive threshold: integral image + local mean (handles uneven lighting)
  - Skew estimation: horizontal projection profile variance
  - Foreground density: popcount for blank page detection
  - Full preprocess_page() pipeline: binarize → skew → density check

For tesseract integration: preprocess with SIMD, then pipe binary image
to tesseract LSTM (which only does character recognition, the fast part).

For our own OCR: binary image → connected components → Base17 fingerprint
per character glyph → codebook lookup = O(1) character recognition.

10 tests: Otsu bimodal, binarize all-white/black/checkerboard,
  density, blank page, text page, skew detection, adaptive vs Otsu.

Data-flow: &[u8] slices (SIMD), owned BinaryImage (write-back), no &mut self.

https://claude.ai/code/session_01ChLvBfpJS8dQhHxRD4pYNp

Author: claude

src/hpc/mod.rs

@@ -224,6 +224,7 @@ pub mod jitson;
 #[cfg(feature = "jit-native")]
 #[allow(missing_docs)]
 pub mod jitson_cranelift;
+pub mod ocr_simd;
 
 #[cfg(test)]
 mod e2e_tests {