import base64 import cv2 import os def get_b64(img_path): if not os.path.exists(img_path): return "" img = cv2.imread(img_path) img = cv2.resize(img, (260, 260)) _, b = cv2.imencode('.jpg', img, [cv2.IMWRITE_JPEG_QUALITY, 90]) return base64.b64encode(b).decode() b64_pre = get_b64("/Volumes/T9/ICML/Part_one_pre_def_rgb/DJI_20250929095743_0311_D.JPG") if not b64_pre: b64_pre = get_b64("/Volumes/T9/ICML/part 2_pre_def_rgb/DJI_20250929093936_0722_D.JPG") b64_post = get_b64("/Volumes/T9/ICML/Post_def_rgb_part1/DJI_20250929124149_0029_D.JPG") html = f"""
QUANTUMHARVEST : HYBRID VAC-QFS INTELLIGENCE
[ I M A G E    R E T R I E V A L    ➔    Q U A N T U M    S P A C E    ➔    Y I E L D    V E R T E X ]
PHASE I: SPATIAL & SPECTRAL PHASE II: VARIATIONAL QUANTUM SUBSET PHASE III: PREDICTIVE SYNTHESIS
PRE-DEFOLIATION
Highly Camouflaged Canopy
POST-DEFOLIATION
Exposed Harvest Topology
Morphological Top-Hat
Tw(I) = I - (I ∘ b)
HSV Masking & CLAHE
M = {{ p | Sp < τs ∧ Vp > τv }}
Contour Distance Transform
D(p) = minq ∈ B d(p, q)
GLCM Textural Matrices
P(i, j | d, θ)
Spectral Indices
ExG = 2G - R - B
Mutual Information Filter Pruning
I(X; Y) = ∑ p(x,y) log [p(x,y) / p(x)p(y)]
Data Encoding
|Φ(x)⟨ = UΦ H⊗n |0⟨
ZZFeatureMap
Parameterized Ansatz
UA(θ) |Φ(x)⟨
RealAmplitudes
Classical Optimizer
minθ L(θ)
COBYLA
Combinatorial Subset Evaluator
maxS ⊂ F, |S|=4 AVQC(S)
Quantum Subset RBF SVM
K(x, x') = exp(-γ ||x-x'||2)
C(6,4) Optimal Manifold
Classification Vertex
ŷ ∈ {{Pre, Post}}
Spatial Yield Density Map
ρ(x,y) = ∑ δ(x-xi, y-yi)
Fig 1: Proposed QuantumHarvest Hybrid VAC-QFS Analytical Pipeline
""" with open("/Volumes/T9/CottonDefoliationApp/difflogic_architecture.html", "w") as f: f.write(html) print("DiffLogic style diagram generated.")